Skip to main content

Your buyers are asking AI. Are you the answer?

Find out with OptimizeCamp →

Changelog

Everything new in SurePrompts — features, fixes, and improvements.

Current version: v2.124.0

v2.124.0
  • ✨FeaturePublished five new ChatGPT photo editing guides: photo editing commands and codes explained, prompts for pictures of yourself, photo enhance prompts that keep the face, passport size photo prompts, and car photo prompts
  • ⚡ImprovementCross-linked the new guides from the photo editing and image prompt collections so each one is reachable in two clicks
v2.123.1
  • 🐛FixRestored Pro templates in the builder search and template picker. Direct template links and saved prompts now wait for Pro templates to load before opening them
v2.123.0
  • 🐛FixFixed two accessibility problems in our emails. The teal used behind white button text, and the grey used for footer and caption text, both fell below the minimum readable contrast. Emails were the last place still using the old colours
  • ⚡ImprovementEmails now use exactly the same colours as the site, so a message and the page it links to look like the same product
v2.122.0
  • ⚡ImprovementCompleted the design refresh across the remaining screens: your dashboard, saved prompts, teams, account, community, sign-in, guides, and the rest of the site. The whole product now uses one consistent colour system
  • ⚡ImprovementReduced the site from sixteen colour families to five. Every colour now means something specific — brand, success, caution, error, or neutral — instead of being decorative
  • 🐛FixFixed several remaining places where status colours did not meet the accessible contrast minimum
v2.121.0
  • ⚡ImprovementApplied the new design to the template builder, JSON builder, Veo3 builder, prompt optimizer, quality score, token counter and comparison tools
  • 🐛FixFixed the prompt quality score indicator using an orange that fell below the accessible contrast minimum. Score and status colours now use one consistent set of success, caution and error colours across the whole product
v2.120.0
  • 🐛FixFixed several call-to-action buttons rendering with invisible label text in dark mode. The button background was fixed white while the label colour followed the theme, so in dark mode the text matched its own background
  • 🐛FixFixed the homepage statistics labels being nearly unreadable in light mode against their dark band
  • ⚡ImprovementTemplate categories now share one consistent badge style, with the category icon carrying the distinction. Category colours were never unique anyway — two categories shared the same one — and the icons work for readers who cannot tell the hues apart
  • ⚡ImprovementApplied the new design to the homepage, template and prompt library browsing, the AI Prompt Generator, billing, and the upgrade dialog. Multi-colour decorative gradients were replaced with a single brand gradient or a flat surface
v2.119.0
  • ⚡ImprovementApplied the new design across the blog index, article pages, series hubs, glossary, author pages, and the comparison, alternative and use-case landing pages. Article and section headings are noticeably larger, and decorative multi-colour background washes have been replaced with a single calm surface
  • 🐛FixFixed white button and table-header text becoming hard to read on teal backgrounds in dark mode. The brand colour is deliberately lighter in dark mode, so those labels now switch to dark text automatically
  • 🐛FixFixed amber helper text in the builder falling below the accessible contrast minimum
  • ⚡ImprovementSocial sharing preview images now use the current brand colours instead of the previous palette
v2.118.0
  • ⚡ImprovementRebuilt how articles read. Body text on every blog post and guide is now set in a serif face at a larger size, on a line length of about 68 characters instead of roughly 90. Long lines were the single biggest thing making articles tiring to read, and this affects all 418 posts
  • ⚡ImprovementHeadings, tables, quotes, callouts and inline links inside articles were restyled to match, and links now carry an underline as well as colour so they stay distinguishable without relying on being able to see the hue
  • ⚡ImprovementRebuilt the site header and footer on the new design. The header background previously did not quite match the page beneath it, and footer link groups now read as groups rather than one long list
  • 🐛FixThe mobile menu button now shows a keyboard focus outline and announces whether the menu is open to screen readers
  • ⚡ImprovementLight and dark colours across the site are now driven by a single set of design tokens rather than being written out by hand on each element. Panels that previously had three slightly different dark backgrounds depending on which page you were on are now consistent
v2.117.0
  • ⚡ImprovementRebuilt the colour system around a single accessible brand teal. The previous brand colour did not meet the WCAG AA contrast minimum against white, despite being used for text in over a hundred places, so it has been deepened slightly. Greys were retuned to match, and every common text-on-background pairing now passes AA in both light and dark mode
  • ⚡ImprovementRedesigned the pricing page: one clear call to action instead of three competing ones, a cleaner plan comparison, and consistent spacing throughout. Checkout behaviour is unchanged
  • ⚡ImprovementLong-form articles, prompt text and interface labels now use three deliberate typefaces rather than one. Prompt and code blocks in particular no longer change appearance between macOS and Windows
  • ⚡ImprovementButtons, cards, inputs, badges and empty states are now built from one shared set of components, so they look and behave consistently as the rest of the site adopts them
  • 🐛FixCards and panels no longer stack a border, a shadow and a rounded corner on top of each other, which made dense pages look boxy
v2.116.0
  • 🐛FixFixed the Table of Contents on blog posts rendering invisibly in dark mode. A global style rule was forcing the page background colour onto it, and onto 78 other panels across the site, so those surfaces disappeared into the page instead of standing out from it
  • 🐛FixRestored the keyboard focus outline on 46 inputs and checkboxes across the builder, JSON builder and advanced options. They were styled to remove the browser default and the replacement indicator was never defined, leaving anyone navigating by keyboard with no way to see where they were
  • 🐛FixFixed 49 buttons, rows and list items whose hover background was stuck on permanently, so hovering them appeared to do nothing
  • 🐛FixThe site now renders in Inter, the typeface it has always loaded. A configuration gap meant the font was downloaded on every page and then discarded in favour of whatever default font your operating system supplies
  • 🐛FixFixed missing dividers and an over-wide text column in the billing and subscription screens
  • 🐛FixFixed dark-mode hover states on the pricing comparison table, the template search and several admin tables, which referenced a shade that did not exist
  • ⚡ImprovementMobile browsers now tint their toolbar and notch area to match the page in both light and dark mode instead of always showing white
  • ⚡ImprovementRebuilt how the light and dark themes are applied. Around 100 lines of override rules that fought the rest of the stylesheet are gone, colour transitions are no longer pinned to 50ms, and the theme still applies before the first paint so there is no flash on load
v2.115.0
  • ⚡ImprovementExpanded the 50 copy-paste ChatGPT prompts guide with a step-by-step section on actually copying and pasting a prompt (desktop, mobile, and the Shift+Enter trick that stops a prompt sending half-written), three new FAQs covering phone copying, accidental sends, and using the same prompts in Gemini or Claude, and a pointer to the image prompt collection for readers after cinematic or aesthetic image prompts
v2.114.0
  • ⚡ImprovementRetired the nine per-model prompt builder landing pages (ChatGPT, Claude, Gemini, Grok, Midjourney, Copilot, Perplexity, DeepSeek, and Llama). Each old "-prompt-builder" URL now permanently redirects to its "-prompt-generator" twin, which already offers the 330+ template builder alongside the AI generator, so bookmarks keep working and search engines see one page per model instead of two competing for the same query. The /ai-prompt-builder doorway page now redirects to the Template Builder at /builder
  • ⚡ImprovementSharpened the search titles and descriptions on the Claude, Gemini, and ChatGPT prompt generator pages and the AI Prompt Generator so they name the models they support, and added a "prompt builder with templates" FAQ to every text-model generator page
v2.113.0
  • ✨FeatureNew guide: 40 ChatGPT photo editing prompts for retouching, background swaps, relighting, old-photo restoration, style transfers, and product listing edits, written for ChatGPT Images 2.0 and its selection brush
  • ✨FeatureNew guide: 60 cool and fun ChatGPT image prompts, from turning a selfie into a boxed action figure or animated character to impossible photos, retro looks, illustration styles, and family projects
  • ✨FeatureNew guide: 35 ChatGPT background prompts covering studio seamless backdrops, headshot and product backgrounds, video-call scenes, website heroes, and phone and desktop wallpapers, with an aspect-ratio table
  • ✨FeatureNew guide: 30 ChatGPT profile picture prompts that turn a selfie into a LinkedIn headshot, social aesthetic, gaming avatar, or illustrated PFP while keeping your likeness
  • ✨FeatureNew guide: 45 AI image prompts for social media posts, organized by platform and format with exact sizes, safe margins, brand style anchors, and batch generation for carousels
  • ⚡ImprovementThe ChatGPT image prompts hub, portrait prompts, image-prompting pillar, and six related guides now link to the new collections
v2.112.0
  • ✨FeatureThe blog index now opens with a "Most Read" list of the 30 guides readers search for most, and the homepage has a "Guides people keep coming back to" section, so the most useful evergreen posts are one click from the front door instead of four
  • ✨FeatureEvery tool page (AI prompt generator, template builder, optimizer, prompt score, pricing, and the general prompt library) now ends with a short "Learn more" block linking the guides, series, and glossary entries that explain how to get the most out of it
  • ⚡ImprovementThe Tools directory gained a "Browse & Learn" section, and the footer now links AI Prompts by Role and Alternatives & Comparisons. Alternatives pages link to the matching prompt generator and model comparisons, comparison pages link to the alternatives, and each model generator lists its comparisons, so none of those families is a dead end any more
  • ⚡ImprovementRelated Articles on every post now reserve one of the three cards for a post that links to the current one, so links flow both ways instead of only outward. Template pages also suggest matching prompt generators in Related Resources
  • ⚡ImprovementGlossary auto-linking now recognises the plain name and the acronym of each term ("mixture of experts", "MoE", "KV cache", "RLHF"), not only the exact glossary title, so 40 of the 41 glossary entries that never received a link from a post now do
  • ⚡ImprovementThe Guides index now lists the Getting Started and Custom Template Builder walkthroughs, which previously linked only to each other
  • 🐛FixInternal links inside posts no longer carry rel="noreferrer", which had been stripping referral data on every in-site click; it is now applied only to links that leave the site
  • 🐛FixPart 5 of AI for Complete Beginners now links to Part 6 instead of saying more parts are on the way
v2.111.0
  • ✨FeatureNew Contact page at /contact, replacing /support (which now redirects there). It asks what your message is about — product support, billing, partnership or media, or something else — so enquiries reach the right place faster
  • 🐛FixContact form submissions now email the team the moment they arrive, with your address set as Reply-To so a reply lands straight in your inbox. Previously messages were only written to a database and nobody was notified, which meant genuine questions could sit unanswered for weeks
  • ⚡ImprovementThe contact form now fills in your name and email reliably when you are signed in, instead of occasionally showing blank fields when the page loaded faster than your account details
  • ⚡ImprovementAdded a hidden spam trap to the contact form so automated submissions are discarded without reaching a person
v2.110.0
  • ⚡ImprovementRefreshed the whole site onto OpenAI's GPT-5.6 generation, which shipped July 9, 2026. GPT-5.6 Sol is now the named flagship across ~80 blog posts, the model landing and comparison pages, the use-case generators, and the prompt generator itself, with Terra and Luna added as the mid and low-cost tiers. GPT-5.5 and GPT-5.4 nano stay listed as still-served previous-generation models
  • 🐛FixCorrected a context-window error the site had carried since July: GPT-5.5's window is 1.05M tokens, not 400K. The 400K figure was Codex's surface cap, not the API model's window. Every comparison built on OpenAI having a smaller window than Gemini or Claude has been reframed — those are now ties, with GPT-5.6 Sol nominally largest at 1.05M
  • 🐛FixUpdated OpenAI pricing to the current rates (Sol $4/$20, Terra $2/$12, Luna $0.20/$1.20 per 1M tokens) and replaced the vague "higher long-context rate" wording with the actual rule: prompts past 272K input tokens bill at 2x input and 1.5x output for the whole request
  • 🐛FixReplaced references to the retired GPT-5 Mini with GPT-5.6 Luna in the cost-sensitive model guides, removed the invented "GPT-5.5 Thinking" product mode (GPT-5.6 folds reasoning into Sol), and corrected the ChatGPT free tier, which now defaults to GPT-5.6 Luna
  • ⚡ImprovementGeneralized Microsoft Copilot's underlying-model claims to "OpenAI's GPT-5 series" rather than asserting a specific point release Microsoft does not publish per tenant
v2.109.0
  • ✨FeaturePublished a five-post agentic-tooling wave: "Codex CLI Prompting Guide" (sandbox vs. approval modes, config precedence, AGENTS.md layering), "Antigravity CLI Prompting Guide" (Google retired Gemini CLI for consumer accounts on June 18, 2026 — migration plus parallel-subagent prompting), "Agent Skills Guide" (SKILL.md, progressive disclosure, portable frontmatter), "AGENTS.md" (the cross-tool instruction file, nesting and precedence), and "AI Browser Agents in 2026" (Comet, Claude for Chrome, ChatGPT's browser agents, and the prompt-injection threat model). All five are interlinked with the coding-agents pillar and the injection-defense canonical
  • ⚡ImprovementAdded Codex CLI and Antigravity CLI to the AI coding agents pillar's tool landscape, and pointed the Claude Code, Cursor-adjacent, computer-use, context-engineering, Gemini, and coding-assistant roundup posts at the new AGENTS.md and Agent Skills guides
  • 🐛FixNew posts name GPT-5.6 Sol as the Codex CLI default model, correcting for OpenAI's July 9, 2026 GPT-5.6 release. The rest of the site still references GPT-5.5 as the OpenAI flagship and is queued for a separate refresh
v2.108.2
  • 🐛FixAI Prompt Generator privacy FAQ and the Prompt Generator page now disclose that AI requests and generated prompts are logged to your account for quality review, replacing the earlier "we do not store" claim
v2.108.1
  • 🐛FixModel landing pages, the Tools index, the Prompt Generator page, llms.txt, and two blog roundups no longer say "no signup required" when linking to the AI Prompt Generator — they now state the free Google sign-in, while template tools stay no-signup
v2.108.0
  • ⚡ImprovementRedesigned the homepage — new two-column hero with the live generator demo, a "tuned for" model strip, a before/after proof section, an on-your-own vs. SurePrompts comparison table, a Pro banner, and a sticky bottom call-to-action. Copy now reflects the Google sign-in flow and current template counts and pricing
  • 🚀EnhancementBoth homepage Pro buttons (banner and pricing card) open the in-page checkout directly, with the monthly/annual toggle carried through
v2.107.0
  • ✨FeatureAI Prompt Generator now requires a free Google sign-in — visitors see an animated sample of what it produces and can sign in with one click; free accounts get 3 generations a month
  • ⚡ImprovementYour typed request and model choice are carried through sign-in, so you pick up exactly where you left off
  • ⚡ImprovementThe anonymous 1-generation-per-month trial has been retired
v2.106.0
  • ⚡ImprovementRe-verified all Token Counter and model-page pricing against each provider's official pricing page (July 2026): corrected Llama 4 Maverick hosted output pricing to $0.80/1M with current host availability, added Grok 4.3's 200K long-context pricing tier, spelled out GPT-5.5's over-272K tier rates, and marked DeepSeek V3.2 as no longer offered on DeepSeek's pricing page.
v2.105.0
  • ⚡ImprovementInterlinked 39 previously hard-to-discover blog posts with 55+ contextual links from topically related guides — every post on the blog now has at least two inbound internal links.
  • ⚡ImprovementContent-health reporting now measures inbound links the way the site actually renders them: related-post sections, series hubs, and landing-page references all count, and the orphan check no longer misreports posts that were already linked.
v2.104.0
  • ⚡ImprovementRefreshed 49 blog posts to the current 2026 model lineup — replacing retired model references (GPT-4o, GPT-3.5, o1/o3, Claude 3, Claude Opus 4.7, DeepSeek R1/V3, Llama 3, Gemini 2.0) with GPT-5.5, Claude Opus 4.8, Gemini 3.1 Pro, DeepSeek V4, and Llama 4 across comparisons, guides, and tutorials, and correcting context-window and pricing facts throughout.
  • ⚡ImprovementRepositioned the o3 strategic-decision prompts and Claude Opus 4.7 prompting guide around current reasoning models, and reframed the voice prompting walkthrough around OpenAI's gpt-realtime model.
v2.103.0
  • 🐛FixCorrected the GPT-5.5 long-context prompts guide to the model's actual 400K-token window (previously claimed 1M tokens), moved it to /blog/gpt-5-long-context-prompts with a permanent redirect from the old URL, and added guidance on when to reach for a true 1M-token model like Claude Opus 4.8 or Gemini 3.1 Pro.
v2.102.0
  • ⚡ImprovementRefreshed all use-case generator pages to the current 2026 model lineup — GPT-5.5, Claude Opus 4.8, Gemini 3.1 Pro, Grok 4.3, DeepSeek V4, and Llama 4 — replacing retired model references (GPT-4o, GPT-3.5, DeepSeek R1/V3, Llama 3, Grok 2/3) and correcting context-window claims across ChatGPT, Claude, Gemini, Grok, DeepSeek, Copilot, and Llama generators.
v2.101.0
  • ⚡ImprovementConsolidated 43 duplicate or misleading content URLs with direct permanent redirects, including generic prompt-generator variants, the duplicate template directory, profession and template-category aliases, duplicate prompt templates, glossary synonyms, and the old library shortcut. Redirect sources were removed from sitemaps, static generation, discovery indexes, and internal links.
  • 🐛FixRepaired broken template references across programmatic landing pages and added build-time integrity tests so generated-page CTAs and recommendation cards cannot silently point to missing templates.
  • ⚡ImprovementRemoved repetitive auto-generated glossary FAQs, corrected the glossary inventory and site-wide template-count claims, and aligned IndexNow and llms.txt output with canonical URLs.
v2.100.0
  • ✨FeaturePublished "How to Get Your Content Cited by ChatGPT and Perplexity" — a practical 2026 GEO guide covering how answer engines choose citations and nine tactics to become the quoted source, interlinked with the GEO, AEO, and AI Overview glossary terms.
  • ⚡ImprovementBlog posts about SEO and AI visibility now end with a contextual OptimizeCamp CTA, and the GEO, AEO, and AI Overview glossary terms gained a Related Tool card linking to it.
  • ⚡ImprovementThe /tools directory now includes a "More From the SurePrompts Team" section linking our sister products: OptimizeCamp, AgentsCamp, and GritShip.
v2.99.0
  • ⚡ImprovementAdded an announcement bar above the site header introducing OptimizeCamp — our sister tool that audits your content for accuracy, authority, and citation readiness so AI assistants like ChatGPT and Perplexity actually reference you.
v2.98.0
  • ✨FeatureExtended the "Which AI Model for X" decision-guide series into three professional verticals: legal work (citation-bound review, discovery routing, damages math), healthcare (deployment-first framing — self-hosted for PHI, documentation and literature picks under clinician review), and marketing content (brand voice, campaigns, trends, volume routing). All three added to the series hub table and interlinked with their vertical prompt guides.
v2.97.0
  • ✨FeaturePublished three more guides, completing the flagship comparison triangle and opening the open-weight cluster: "GPT-5.5 vs Claude Opus 4.8" and "Gemini 3.1 Pro vs Claude Opus 4.8" (each interlinked with the GPT-5.5 vs Gemini face-off), plus "Llama vs Mistral vs DeepSeek" — the open-weight model decision guide covering ecosystem, EU compliance, self-hosting economics, and fine-tuning.
v2.96.0
  • ✨FeaturePublished three new guides: "GPT-5.5 vs Gemini 3.1 Pro" — the flagship model face-off, "Grok 4.3 vs DeepSeek V4" — the challenger comparison (real-time data vs open-weight economics), and "How to Stop AI Hallucinations" — nine prompting fixes with a practical verification workflow. All three are interlinked with the existing comparison cluster and glossary.
v2.95.0
  • ⚡ImprovementInterlinked the 20-part "prompt patterns" blog series into a proper topic cluster. Each guide now points to three closely-related sibling guides (e.g. the API-documentation patterns link to bug-report, code-review, and technical-docs patterns), so every post in the set is well-connected instead of nearly isolated. Readers get more relevant "Related" suggestions, and the cluster reads as a single authoritative body of work to search engines.
v2.94.0
  • ⚡ImprovementInternal-linking clean-up across the blog: repointed 30 in-article links that pointed at persona pages which only redirect search engines onward (e.g. /prompts-for/marketers) straight to their canonical prompt-generator landing pages. This stops a hop of link equity from leaking and consolidates ranking signals on the richer destination pages.
v2.93.0
  • ⚡ImprovementTechnical SEO clean-up from a full-site crawl audit: gave six near-duplicate prompt templates distinct titles so they no longer compete for the same search results, capped meta descriptions at the search-snippet length (with clean word boundaries) so they stop truncating mid-sentence, and gave each blog archive page a unique description. Stopped the sitemap from advertising persona pages that already point search engines elsewhere, fixed a stray Runway Gen-3 prompt-library link (plus a redirect for any shared copies), and removed conflicting "canonical to homepage" tags from non-indexed pages.
v2.92.0
  • 🐛FixFixed completed Pro purchases not being recorded as conversions in analytics. Checkout completions are now reported server-side from the Freemius license webhook — the authoritative "payment received" signal — via the GA4 Measurement Protocol, so revenue is captured reliably even when checkout finishes through a redirect or the buyer hasn't granted analytics consent. Signups are also now attributed to their sign-in method.
v2.91.0
  • ⚡ImprovementSwept the model-comparison and per-model SEO landing pages (ChatGPT/Claude/Gemini/DeepSeek/Llama/Grok comparison pages, generator pages, and prompt-builder pages) plus five evergreen guides onto the current 2026 lineup — GPT-5.5, Claude Opus 4.8, Gemini 3.1 Pro, DeepSeek V4, Llama 4 Maverick, Grok 4.3 — with verified context windows and pricing. Retired GPT-4o/GPT-4, Claude 3.5/3 Opus, Gemini 1.5/2.0, DeepSeek V3/R1, o1/o3, and Llama 3 from anywhere they were presented as current. Regenerated the llms.txt machine-readable indexes to match.
v2.90.0
  • ⚡ImprovementRefreshed the last three model-stale "which AI model for…" decision guides — long-context document analysis, creative writing, and math & quantitative reasoning — to the current 2026 lineup (GPT-5.5, Claude Opus 4.8, Gemini 3.1 Pro, DeepSeek V4), retiring Opus 4.7, bare GPT-5, o3, and the old Gemini flagship. The long-context guide was reframed now that no model exceeds a 1M-token window: the comparison shifts from raw window size to multimodality, deep-retrieval accuracy, reasoning, and cost, with verified context windows throughout. All three now align with the model-selection hub.
v2.89.0
  • ⚡ImprovementRefreshed the model data powering the AI Prompt Generator, the model directory (/models), and the Token Counter & Cost Calculator to the current 2026 flagships — Gemini 3.1 Pro and DeepSeek V4 are now the Google and DeepSeek defaults, with pricing and context windows verified against each provider's official page. Generated prompts now name the current models instead of the retired Gemini 2.5 Pro / DeepSeek V3.2 labels, and both new models are selectable in the cost calculator.
v2.88.0
  • ✨FeaturePublished six new "which AI model for…" decision guides — data analysis & spreadsheets, research & multi-document synthesis, real-time research & current events, customer support & chatbots, translation & multilingual, and private/self-hosted workloads. Each gives a model-by-model comparison matrix, sub-segment recommendations, and a copy-paste sample prompt for the recommended winner, using the current 2026 lineup (GPT-5.5, Claude Opus 4.8, Gemini 3.1 Pro, Grok 4.3, DeepSeek V4).
  • ⚡ImprovementRefreshed the "which AI model for coding" guide to the current 2026 lineup (GPT-5.5 / Claude Opus 4.8 / Gemini 3.1 Pro / DeepSeek V4) and wired the new guides into the model-selection hub for easier discovery.
v2.87.0
  • ⚡ImprovementRefreshed our model-selection and prompt-collection guides to the current 2026 flagship lineup — the "which AI model for vision, charts & PDFs" and "which AI model for reliable agents" comparisons, the 1M-context codebase-review and agent/tool-use prompt collections, and the long-context prompts guide now reference GPT-5.5, Claude Opus 4.8, and Gemini 3.1 Pro instead of retired models, with every copy-paste prompt preserved. Also repaired a broken internal link and added contextual cross-links from the model-selection guide so these guides are easier to find.
v2.86.0
  • ⚡ImprovementUpgraded the AI Prompt Generator to a GPT-5-class engine (from GPT-4o mini) and refreshed its model-specific tuning to the current 2026 lineup — prompts generated for ChatGPT, Claude, Gemini, Grok, and DeepSeek now reference the current flagship models (GPT-5.5, Claude Opus 4.8, Gemini 2.5 Pro, Grok 4.3, DeepSeek V3.2) instead of retired ones, for higher-quality, better-targeted output on every generation.
v2.85.0
  • 🐛FixRefreshed the model-specific prompt collections to the current lineup — best ChatGPT, DeepSeek, Gemini, GPT-5, Mistral, and Copilot prompts now reference GPT-5.5 (and DeepSeek V4) instead of the retired GPT-4o. Reframed the o3 collection as “Best Reasoning-Model Prompts” around GPT-5.5 high reasoning effort, and updated the Opus 4.7 collection to Opus 4.8 — every copy-paste prompt preserved.
v2.84.0
  • ⚡ImprovementSite-wide internal linking audit: repaired 20 broken links left over from past post renames/consolidation, and added ~120 natural contextual links so related guides connect better and our most useful pages — the AI Models comparison, the model-selection hub, and high-traffic how-to and comparison guides — are easier to discover.
v2.83.0
  • 🐛FixUpdated the best Perplexity prompts guide to the current model roster — Perplexity Pro now offers GPT-5.5, Claude Opus 4.8 and Sonnet 4.6, Gemini 3.1 Pro, and Grok 4.3 (replacing the older GPT-5.2 / Claude 4.6 references).
v2.82.0
  • 🐛FixRefreshed the two Grok guides (best Grok prompts, how to use Grok) to the current lineup — corrected a stale "Grok 3 / 131K context" stat to Grok 4.3 / 1M tokens and replaced retired GPT-4o comparison references with current models.
v2.81.0
  • ⚡ImprovementSharpened the search titles and meta descriptions on three high-impression guides (best Grok prompts, best Perplexity prompts, how to use Grok) — tightened so they no longer truncate in search results and lead with the copy-paste / cited-research value — to lift click-through on the impressions they already earn.
v2.80.0
  • ⚡ImprovementRefreshed the 20-prompt same-prompt comparison guide to the current frontier trio — GPT-5.5 vs. Claude Opus 4.8 vs. Gemini 3.1 Pro — replacing the retired o3 / GPT-5 / Opus 4.7 lineup. All 20 copy-paste prompts are unchanged; the per-task verdicts, model-selection rubric, and FAQs now reflect each current model’s real strengths.
v2.79.0
  • ⚡ImprovementRefreshed seven model-related guides that had fallen behind the current AI lineup — both reasoning-model guides, How to Use DeepSeek, ChatGPT vs Claude, 22 Best AI Tools, Which AI Model Should You Use, and 9 AI Models Compared. Replaced every retired model (o3/o4-mini, GPT-4o, Claude Opus 4.6, DeepSeek R1/V3, older Grok) with the verified June 2026 lineup (GPT-5.5, Claude Opus 4.8 / Fable 5 / Sonnet 4.6, Gemini 3.1 Pro, DeepSeek V4) — including updated titles, pricing, and benchmark framing, with primary sources replacing low-credibility citations.
v2.78.0
  • ⚡ImprovementFully refreshed our flagship AI model comparison guide (Best AI Model in 2026) with the current lineup — GPT-5.5, Claude Opus 4.8 and Fable 5, the Gemini 3 family, Grok 4.3, and DeepSeek V4 — plus verified June 2026 API pricing, context windows, and consumer plan tiers. Removed stale, retired models (GPT-4o, o3, older Grok) and replaced low-credibility benchmark citations with primary sources.
  • 🐛FixCorrected the Gemini 2.5 Flash price in the token counter and model pages to $0.30/$2.50 per million tokens (the previous figure was actually Flash-Lite’s), added a Gemini 2.5 Flash-Lite row, and set Claude Opus 4.8 and Sonnet 4.6 to their 1M-token context window.
v2.77.0
  • ✨FeatureNew Research hub (/research) that gathers our original State of AI Prompting data studies in one place — key findings, links to each report, and downloadable open datasets (CC BY 4.0) anyone can cite.
v2.76.0
  • ✨FeatureNew AI Model Comparison hub (/models) plus a reference page for each major model — context window, API pricing, modalities, capabilities, and prompting tips for GPT-5.5, Claude, Gemini, Grok, DeepSeek, Llama and more. All figures web-verified against provider docs.
  • 🐛FixCorrected context-window figures in the cost calculator: Grok 4.3 and Llama 4 Maverick are 1M tokens, and DeepSeek V3.2 is ~160K.
v2.75.0
  • ✨FeatureNew free tool: Token Counter & LLM Cost Calculator (/token-counter). Paste any text to count tokens with OpenAI's o200k tokenizer and instantly compare API cost across GPT-5, Claude, Gemini, Grok, DeepSeek and more — all in your browser, nothing uploaded.
v2.74.0
  • ⚡ImprovementConsolidated a handful of overlapping blog guides into their stronger canonical versions (old links redirect automatically) and refocused several others on distinct topics, so each guide is clearer about what it covers.
  • ⚡ImprovementAdded contextual links throughout the blog so related guides connect to each other, making it easier to navigate between topics.
v2.73.0
  • ✨FeatureNew data study — "People Write Better Prompts for Claude Than for ChatGPT" breaks down our State of AI Prompting dataset by AI model, showing how prompt quality varies across ChatGPT, Claude, Gemini, Copilot, and more.
  • ✨FeatureNew data study — "The Anatomy of a Failing AI Prompt" traces where the average prompt loses 79 of its 100 points, and the three fixes that recover more than half.
  • 🚀EnhancementPublished the State of AI Prompting research as downloadable open datasets (CC BY 4.0) so anyone can cite and build on the numbers.
v2.72.0
  • ⚡ImprovementThe blog index is now paginated (18 articles per page) with Previous/Next navigation, so the page loads faster and is easier to browse instead of rendering every article at once.
v2.71.0
  • ⚡ImprovementRefreshed and expanded our most-read AI prompt guides (ChatGPT, Gemini, Grok, DeepSeek, Perplexity, image prompts, designers, e-commerce, and job interviews) with new workflow and comparison sections and links to related tools and guides.
v2.70.0
  • ✨FeatureIntroducing AgentsCamp (agentscamp.com), our new sister site — a curated library of 200+ drop-in resources (agents, skills, slash commands, tools, and guides) for building with AI coding agents like Claude and Claude Code. Read the full announcement on our blog.
  • ⚡ImprovementAdded an AgentsCamp link to the site footer so you can explore our sister site from any page.
v2.69.0
  • ✨FeatureNew free "Learn" hub at /learn — a single landing page that showcases all of our step-by-step AI guide series in one place, so visitors can pick a learning path and start for free.
  • ⚡ImprovementStreamlined the main navigation: added Learn to the header, and moved the Prompt Tools menu and the Community link into the footer to keep the top nav focused.
v2.68.0
  • ✨FeatureCompleted the "AI for Complete Beginners" series — Parts 6 through 10 are now live, covering AI for everyday life, AI at work, when AI gets it wrong, which AI tool to use, and building your prompt library.
  • ✨FeatureLaunched six new step-by-step guide series, each with its own hub under /blog/series: "Score Your Prompt: The 7-Day Challenge", "Prompting Pro in 21 Days", "Your First AI Agent", "Career-Proof", "Visuals That Sell: AI Image and Video for Non-Designers", and "AI for the One-Person Business" — 49 new guides in total.
  • ⚡ImprovementAdded series hub pages with ItemList and breadcrumb structured data so each multi-part series is easy to navigate and eligible for richer search results.
v2.67.0
  • ✨FeatureLaunched "AI for Complete Beginners" — a free, plain-English series for anyone new to AI. The first three parts are live (What Is AI, Really?; Your First AI Conversation; What Is a Prompt?), with a series hub at /blog/series/ai-for-beginners and more parts on the way.
  • ✨FeatureEach new guide in the beginners series is shared with subscribers by email as free educational content, with one-click unsubscribe — so you can follow along without having to check back.
v2.66.0
  • ✨FeatureEvery blog article now has a Frequently Asked Questions section backed by FAQ structured data. We added 4-6 question-and-answer pairs to 216 posts, bringing the entire 329-article library to full coverage — making each post eligible for rich results and far more likely to be quoted by AI answer engines like Google AI Overviews, Perplexity, and ChatGPT.
  • 🐛FixFixed broken social-share preview cards on every "alternatives" comparison page — they referenced image files that did not exist, so links shared on X, Slack, and LinkedIn showed a broken card. They now render the correct branded preview.
  • 🐛FixRemoved inaccurate 5-star ratings from community template pages. Upvotes were being reported to Google as star reviews; the structured data now represents engagement honestly, avoiding a review-snippet policy issue.
  • 🐛FixCorrected the organization logo used for Google's knowledge panel (it pointed at a missing file), and cleaned up page titles that showed the brand name twice ("… | SurePrompts | SurePrompts") on glossary, template, and tool pages.
  • ⚡ImprovementArticles and guides now carry real author bylines that link to dedicated author pages, backed by proper Person/author structured data — strengthening trust and authorship signals (E-E-A-T) across the content library.
  • ⚡ImprovementModel, comparison, use-case, persona, template-category, and "prompts for" landing pages now generate their own branded social-share preview images instead of sharing one generic card.
  • ⚡ImprovementProgrammatic pages now show an honest "Last updated" date, and the RSS/JSON feeds and sitemap report real, stable content dates instead of changing on every deploy.
  • ⚡ImprovementConsolidated duplicate category and role pages onto a single canonical version each, so overlapping pages stop competing with one another in search.
  • 🚀EnhancementTemplate detail pages now surface more of each template's details as readable content, with an automatic quality check that keeps thin or near-duplicate pages out of the search index.
  • 🚀EnhancementAdded answer-first summaries, machine-readable "speakable" statistics, and dataset structured data (including for the State of AI Prompting 2026 study) to improve how AI answer engines read and cite the site.
v2.65.0
  • ✨FeatureNew free Prompt Quality Score tool at /prompt-score — paste any prompt and get an instant 0-100 score across 8 categories with specific suggestions to improve it, plus an embeddable score badge you can drop on your own site.
  • ⚡ImprovementEvery glossary term now has a Frequently Asked Questions section backed by FAQ structured data — clearer for readers and eligible for rich results and AI answer engines.
  • ⚡ImprovementSmarter internal links: related-article suggestions are now reciprocal, and blog posts automatically link the relevant templates and guides, making related content easier to find.
  • ⚡ImprovementConsolidated several overlapping blog posts and added redirects so each topic has one definitive page, strengthening search rankings.
v2.64.0
  • 🐛FixFixed broken canonical URLs on the AI Prompt Builder and Prompt Generator landing pages — a code typo was sending search engines a malformed canonical, which could split ranking signals. Both now point to their correct addresses.
  • ⚡ImprovementTemplate, glossary, and community pages now generate their own branded social-share preview images, so links shared on X, LinkedIn, and Slack show a custom card with the item's title instead of a generic image.
  • ⚡ImprovementPolicy pages (privacy, cookies, terms) now include proper Open Graph tags for cleaner link previews when shared.
  • ⚡ImprovementThe pricing page FAQ shown to visitors now exactly matches its structured data, making it eligible for rich FAQ results in Google.
  • ⚡ImprovementSurePrompts is now installable as an app on supported devices via a web app manifest.
  • ⚡ImprovementFaster page loads: the analytics consent script no longer blocks initial rendering.
  • ⚡ImprovementTemplate cards across category, model, and use-case pages now open each template's dedicated detail page (with full description and a one-click "open in builder" action) instead of jumping straight into the tool — clearer browsing and better-indexed template pages.
v2.63.0
  • ✨FeatureBrand-new homepage built around a single provocative idea — most people only tap ~10% of what AI can do, and the other 90% is unlocked by the prompt. Features a bold kinetic hero, an animated "capacity meter" that visualizes a weak prompt (31/100) vs. a SurePrompts-engineered one (94/100), a live demo of the generator turning plain English into an expert prompt, a bento feature grid, a Pro pricing preview with a monthly/annual toggle, and an expandable FAQ.
v2.62.0
  • 🐛FixThe Veo 3 prompt builder was showing the free-limit upgrade wall to every logged-out visitor on first load, blocking the form entirely — anonymous visitors can now use it as intended.
  • 🐛FixCreating a public share link ("Create Public Link") now works; it was failing for everyone due to a missing request-security token.
  • 🐛FixDay-3 and day-7 onboarding emails now send reliably — the scheduled job was being invoked with a method the route did not accept, so it silently never ran.
  • ⚡ImprovementAccount-security hardening: closed several access-control gaps across community templates, teams, shared prompts, AI usage limits, and the admin area, and tightened API rate-limiting against abuse. No action needed by users.
  • ⚡ImprovementAI generation credits are no longer consumed when a generation fails partway — failed attempts are now refunded.
  • 🐛FixThe prompt optimizer free-tier daily limit now correctly gates the improved-prompt output (previously only the score was withheld).
  • 🐛FixResolved duplicate-URL and metadata mismatches on the ChatGPT and Midjourney prompt-generator landing pages, and made the sitemap cacheable for faster, lighter crawls.
  • 🐛FixCode samples and multi-value fields containing "$" sequences now render correctly in blog/guide content and generated prompts.
v2.61.0
  • ✨FeatureThree new copy-paste prompt collections in our highest-converting format (the proven /blog/chatgpt-image-prompts-2026 template). /blog/chatgpt-prompts-copy-paste — 50 ready-to-paste ChatGPT text prompts across writing, email, business, coding, analysis, learning, and productivity, targeting the high-intent "chatgpt prompt copy paste" query cluster (~2,000 monthly impressions currently leaking to the image post). /blog/grok-imagine-prompts-copy-paste — 45 copy-paste image and video prompts for Grok Imagine, capturing untapped "grok video generation prompt examples" demand. /blog/business-ai-prompts-copy-paste — 40 business prompt packs for sales, marketing/competitive analysis, finance, strategy, and operations. Each post carries FAQPage schema and contextual links to the matching prompt generators.
v2.60.0
  • ✨FeatureNew /tools hub — a single directory linking every prompt generator and builder (by AI model, use case, and profession) plus model comparisons. Linked from the footer so all 160+ landing pages now have an authoritative internal link instead of being reachable only from each other.
  • ⚡ImprovementBlog posts now surface a contextual "Free Tool" link to the matching prompt generator in their Related Resources section, driven by each generator's curated relatedBlogSlugs (with keyword-overlap as a fallback). This strengthens internal linking from indexed blog content to the generator pages to help them get crawled and indexed.
v2.59.2
  • ⚡ImprovementResolved a "Duplicate without user-selected canonical" issue in Google Search Console affecting /builder?template= URLs. Those pages declared both a canonical to /builder and a noindex tag — contradictory signals that caused Google to ignore the canonical. Removed the noindex so the canonical alone consolidates the parameter variants onto /builder, the recommended handling for parameter duplicates.
v2.59.1
  • 🐛FixFixed the three 404s flagged in Google Search Console coverage. The /prompt-libraries and /prompt-libraries/general pages had an escaped template literal (\${siteUrl}) in their canonical/Open Graph URLs, which Google resolved against the site root and crawled as broken https://sureprompts.com/${siteUrl}/... URLs — now interpolated correctly. The blog footer CTA also linked to /prompts-for/sales (and several other non-existent persona slugs: designers, small-business, presentations, compliance, therapists); these now point to the closest existing prompt collections, eliminating the broken internal links that surfaced the 404.
v2.59.0
  • ✨FeatureFive new posts extending proven content patterns. Three anchor-split image-prompt listicles capture demand from sections of /blog/chatgpt-image-prompts-2026 that Google indexed as separate URLs with high impressions and zero clicks: /blog/chatgpt-portrait-prompts (40 prompts), /blog/chatgpt-product-photography-prompts (35 prompts), /blog/chatgpt-food-photography-prompts (30 prompts). /blog/best-mistral-prompts-2026 fills the obvious gap in our "best X prompts 2026" pattern (30 prompts across multilingual writing, Codestral coding, EU compliance, research, structured outputs, creative). /blog/ai-subscription-plans-compared-2026 captures the "ChatGPT Plus vs Claude Pro vs Gemini Advanced vs Perplexity Pro" tier-comparison demand surfacing in GSC.
v2.58.0
  • ✨FeaturePage-2 → page-1 depth upgrade on eight AI model comparison posts (Grok vs ChatGPT, ChatGPT vs Claude, Perplexity vs ChatGPT, Copilot vs ChatGPT, Claude vs Gemini, GPT-4o vs Claude Sonnet, DeepSeek vs ChatGPT, Gemini vs ChatGPT). Each post now opens with a Quick Verdict block (the format AI Overviews scrape), carries a "Pricing verified 2026-05-28" note above the comparison table, ends with a Common Questions section answering 7 long-tail GSC query variants, ships an FAQPage-schema faqs array in frontmatter, and links contextually to /blog/complete-guide-ai-models-2026. ~8,500 words added across the wave.
v2.57.0
  • ⚡ImprovementCTR rescue on six high-impression, low-CTR pages flagged by GSC (1.3M combined impressions, sub-1% click-through). Rewrote titles and meta descriptions on /blog/complete-guide-ai-models-2026, /blog/flux-pro-prompting-guide, /blog/how-to-write-ai-image-prompts, /blog/best-ai-tools-2026, /blog/midjourney-v7-prompting-guide, and /blog/ai-reasoning-models-prompting-complete-guide-2026 to the proven NUMBER + OUTCOME + YEAR pattern (e.g. "22 Best AI Tools in 2026 (Tested & Ranked by Category)"). Counts and formulas verified against actual content. updatedAt bumped on all six.
v2.56.0
  • ✨FeatureSix new "best ----" listicles covering audience-specific and category-specific buyer queries. Audience guides: Best AI Tools for Marketers, Best AI Tools for Students (with an honest academic-integrity section), and Best AI Tools for Content Creators — each with a curated 10-tool stack and example stacks by team size or content format. Category deep-dives: Best AI Image Generators (8 tools — Midjourney, DALL-E 3, Nano Banana, Adobe Firefly, Flux, Stable Diffusion, Ideogram, Recraft), Best AI Video Generators (8 tools — Sora 2, Veo 3, Runway, Pika, Kling, Luma, Hailuo, Hunyuan), and Best AI Coding Assistants (8 tools — Claude Code, Cursor, Copilot, Windsurf, Aider, Cline, v0, Devin). All six emit ComparisonTable JSON-LD via frontmatter and follow the honesty rule.
v2.55.0
  • ✨FeatureThree new "best ----" listicles targeting high-intent buyer-research queries: Best AI Prompt Generators in 2026 (8 tools compared, distinct from prompt libraries), Best Prompt Engineering Tools in 2026 (9 tools across the full workflow stack — generation, observability, eval, management, optimization), and 10 Best Free AI Prompt Tools in 2026 (free-tier focus across generators, libraries, open-source tools, and free chatbots). Each post emits ComparisonTable JSON-LD via frontmatter and follows the honesty rule — qualitative ratings, no fabricated stats or invented features.
v2.54.0
  • ✨Feature7 new "Which AI Model for X" decision-matrix guides covering coding, long-context document analysis, building reliable agents, vision/chart/PDF understanding, creative writing, cost-sensitive workloads, and math/quantitative reasoning. Each post carries comparisonTable frontmatter that emits ComparisonTable JSON-LD for richer search results. Honesty rule: capability ratings are qualitative (Best-in-class / Strong / Adequate / Trailing), not fabricated benchmark percentages.
  • ✨FeatureSix new glossary entries (205 → 211): tau-bench, osworld, ruler-benchmark, model-routing, cost-per-task, function-calling-accuracy. The auto-linker now highlights these inside the decision-matrix posts and any future content that references them.
v2.53.3
  • 🐛FixRepaired 6 broken internal links across blog content uncovered by a full audit. Fixed: the "ai-prompts-for-research" relatedPosts entry on perplexity-vs-chatgpt-2026 (now points to ai-prompts-investment-research), a 404 /prompts/google-gemini link on how-to-use-gemini (now routes to /prompt-libraries/general), three /glossary/retrieval-augmented-generation references that should have been /glossary/rag, and one /glossary/multimodal reference that should have been /glossary/multimodal-prompting
v2.53.2
  • 🐛FixFixed broken "Related Guide" links on ~20 glossary pages (including /glossary/context-engineering). The curated related-content slug now resolves to a guide first and falls back to a blog post, so cards like context-engineering, prompt-caching, cursor-ai-prompting, and the multi-agent terms link to their actual pillar/post URLs instead of 404ing. The card also now shows the destination article title instead of a generic teaser
v2.53.1
  • ⚡ImprovementTrimmed indexable surface for Google — removed thin UUID-based prompt-library detail pages and /builder?template= variants from the sitemap, marked both noindex with canonicals consolidating signals onto the stable hubs (/prompt-libraries/{model}, /prompts/{category}/{id}, /builder). The cleanup should restore enough domain-trust signal for previously-crawled-but-not-indexed blog posts and landing pages to enter the index on the next crawl
  • ⚡ImprovementGlossary term pages now include a structured "How it works" walkthrough, a "Not to be confused with" disambiguation block, and an inline "Origin" reference. Backfilled the new sections on five previously not-indexed terms: ColBERT, prefix-tuning, self-debug, self-ask prompting, and vision-language models
  • ⚡ImprovementGeneric "Put this into practice" CTA on glossary pages now only renders when the term has no related guide and no cross-content links — removing duplicated boilerplate from the ~200 pages that already surface richer engagement options
v2.53.0
  • ✨FeatureLifecycle email drips for new free users — added day-3 activation nudge and day-7 Pro pitch, both sent via a daily Vercel Cron sweep. Both respect the existing unsubscribe / suppression list and skip Pro and admin users. Idempotency is tracked per-drip via flags on auth.users.user_metadata so the same user can never receive the same drip twice
  • ✨FeatureNear-limit upgrade email — free users who land on exactly 1 AI generation remaining for the month now get a one-time-per-month email reminder. Triggered server-side from /api/generate-prompt the moment the threshold is crossed, with a month-scoped idempotency flag so the next month re-arms naturally
  • ✨FeatureIn-app "1 generation left" banner on the AI Prompt Generator — softer warm nudge that renders once per session before the hard paywall at zero. Links to /pricing with utm_source=in_app_nudge and tracks upgrade_nudge_shown / upgrade_nudge_clicked in GA4
  • ⚡ImprovementAdded signup_completed and first_generation_completed GA4 events to make the Pro conversion funnel measurable end-to-end (signup → activation → paywall → checkout). Both events are de-duplicated per browser so they fire exactly once per user
v2.52.2
  • 🐛FixCommunity gallery cards are now clickable — both the title and a new "View" button open the full prompt page. Previously there was no way to drill into a published prompt from the grid
  • 🐛FixRemoved the "Open this template in Builder" CTA from community prompt pages — many published prompts are AI-generated and don't have a corresponding builder template, making the button misleading
  • 🐛FixResolved a React key collision in the admin template-usage table when multiple template entries shared the same template_id
v2.52.1
  • 🐛FixFixed Pro subscription not unlocking the AI Prompt Generator monthly cap — affected users were stuck at the 3/month free limit despite an active subscription because a backend lookup referenced a missing column and silently fell back to free-tier defaults. Pro accounts now correctly get 200 generations/month
  • 🐛FixAI Prompt Generator usage counter no longer flickers on page load — eliminated a race where a stale localStorage count was shown briefly before the server sync, and a transient "0 of 200" flash for Pro users while auth was still loading
v2.52.0
  • ✨FeatureCommunity Publishing — any logged-in user can publish a saved prompt to the public Community gallery. Each published prompt gets its own SEO-optimized permalink at /community/[slug] with full Open Graph, Twitter Card, and SoftwareApplication + BreadcrumbList JSON-LD. Permalinks ping IndexNow on publish and are included in the sitemap. Publish from the My Prompts page or the Builder via a one-click modal (title, description, category, tags). Auto-takedown after 3 reports with admin review at /admin/community/reports
  • 🐛FixAdded robots noindex to /shared/[slug] pages and fixed an escaped template literal in the canonical URL — eliminates duplicate-content competition with the new /community/[slug] permalinks
v2.51.0
  • ✨Feature23 new copy-paste prompt roundups across the 2026 frontier-model lineup — ~155,000 words of paste-ready prompts. Text models: 50 GPT-5 prompts, 30 GPT-5 long-context (1M-token) prompts, 25-prompt GPT-5 Custom Instructions library, 40 GPT-5 coding prompts, 50 Claude Opus 4.7 prompts, 30 Opus 4.7 agent/tool-use prompts, 25 Opus 4.7 1M-context codebase-review prompts, 30-persona Claude Projects library, 50 o3 prompts (math, code, planning, research), 30 o3 prompts for strategic decisions, and a 20-prompt o3-vs-GPT-5-vs-Opus-4.7 head-to-head. Image models: 50 Midjourney v7 prompts, 30 MJ v7 logo & brand-identity prompts, 30 MJ v7 cinematic prompts, 50 Nano Banana (Gemini 2.5 image) prompts, 30 Nano Banana product-photography prompts, 20-prompt Nano Banana vs MJ v7 head-to-head. Video models: 50 Sora 2 prompts, 30 Sora 2 cinematic prompts, 30 Sora 2 product-video prompts, 50 Veo 3 prompts (with native audio direction), 30 Veo 3 prompts for YouTube Shorts/TikTok, 30 Veo 3 prompts for brand & product videos. Each post follows the established copy-paste roundup format with category headers, fenced code blocks, power tips, and before/after comparison
v2.50.1
  • 🐛FixResolved a Freemius webhook regression that prevented Pro entitlement from activating after checkout in some cases. Affected upgrades have been replayed and Pro access now activates immediately
v2.50.0
  • ⚡ImprovementFull-corpus exposure to AI engines — llms.txt expanded from 61 → 350 lines (now indexes the full 205-term glossary, 12 pillars, 9 canonical posts, 11 template categories, 4 named frameworks) and llms-full.txt expanded from 6,399 → 8,698 lines (now includes 262 tutorial summaries + 118 templates vs. the previous 21 entries — was hiding ~95% of the content footprint from ChatGPT/Claude/Perplexity/Gemini crawlers). Both files now auto-regenerate from real data via the prebuild script
  • ⚡ImprovementDirect-answer summary frontmatter backfilled on 105 additional blog posts. The full corpus of ~280 posts now renders a TL;DR card on page load and exposes Speakable schema for voice/AEO. Also backfilled updatedAt on 11 of 12 pillar canonicals so dateModified renders on every pillar
  • ✨FeaturePer-post OG image auto-generation for every blog post via Next.js opengraph-image convention — 1200×630 PNG with category pill, post title, and brand wordmark on a dark slate gradient. Replaces the single generic share card that all 280+ posts previously shared
  • ✨FeatureJSON Feed v1.1 endpoint at /feed.json mirroring RSS with date_modified, summary, content_html, and author. Discoverable via metadata.alternates.types — preferred by Perplexity/Gemini over XML for ingestion
  • ✨FeatureHowTo schema rendering on step-by-step tutorial posts. Initial coverage: how-to-use-ai-prompt-generator, sureprompts-complete-tutorial, sora2-prompts-guide, ai-prompt-generator-guide, ai-model-selection-guide. Powers procedural rich snippets in Google + improves citation odds on procedural-reasoning queries
  • ✨FeatureComparisonTable JSON-LD schema rendering pipeline for "X vs Y" comparison posts (opt-in via comparisonTable frontmatter). Component now supports both the existing A-vs-B variant used on AI-model landing pages and a new generic multi-item variant for blog comparisons
  • ⚡ImprovementVisible "Updated [date]" line rendered on blog posts when updatedAt differs from publishedAt — surfaces freshness signal as page text, not just schema (LLMs cite recent content ~40% more)
  • ⚡ImprovementBreadcrumbList JSON-LD schema added to /prompt-libraries/[model] pages. Organization schema in root layout now exposes contactPoint (customer support) so AI engines can surface support routes
  • 🐛FixGlossary auto-linker now supports Grok and xAI with case-sensitive scoping. Capitalized "Grok" links to the glossary entry; the English verb "grok" does not. "XAI"/"xAI" link; lowercase "xai" does not. Both were previously excluded due to false-positive risk, leaving the entries orphaned. 9 new unit tests added (29 total, all passing)
  • ⚡Improvement101 over-long glossary entries restructured to lead with a ≤60-word direct-answer definition; longer nuance/examples moved to a new elaboration paragraph. Speakable selector still targets only the crisp definition so AEO citation surface stays focused
  • 🐛FixAuth pages (login, signup, callback) now correctly noindex via metadata robots config; admin/account/library/my-prompts/support pages noindex+nofollow for defense-in-depth (robots.txt was already disallowing). Added metadata exports to 15 internal pages that previously had none
  • ⚡ImprovementHome page title shortened to 61 chars (was 68, getting truncated in SERPs); Pricing page title expanded to 57 chars (was 30, wasting SERP real estate)
v2.49.0
  • ✨Feature4 new tutorials targeting the AI memory architectures trending cluster — third wave of the day. Deepens the existing high-level ai-memory-systems-guide with framework-specific walkthroughs. (1) "Letta (MemGPT) Walkthrough" — open-source agent framework where the agent self-manages memory via tool calls (memory blocks, archival memory, OS-inspired hierarchy). (2) "mem0 Implementation Guide" — open-source memory layer (add/search/update/delete primitives, multi-level scoping, optional graph memory). (3) "Agent Memory Architectures Compared (2026)" — reference comparing the 5 main camps (provider-managed, Letta, mem0, vector RAG, custom) with decision framework and migration paths. (4) "Episodic vs Semantic Memory for AI Agents" — cognitive-science framing (Tulving 1972) mapped to agent design decisions, with procedural memory as the third pillar. ~12,200 words combined
  • 🚀Enhancement10 new glossary entries supporting the AI memory wave: episodic-memory, semantic-memory, procedural-memory, working-memory, memory-block, memory-recall, vector-memory, conversation-memory, mem0, letta. Glossary now at 205 entries
v2.48.0
  • ✨Feature4 new tutorials targeting the agent-frameworks 2026 trending cluster — zero existing site coverage, wide-open SERPs, developer-heavy traffic. (1) "LangGraph Prompting Guide" — LangChain's graph-based stateful multi-agent framework (StateGraph, MessagesState, conditional edges, interrupts, checkpointers). (2) "CrewAI Prompting Guide" — role/goal/backstory multi-agent framework (sequential vs hierarchical process, delegation, memory). (3) "OpenAI Agents SDK Prompting Guide" — OpenAI's production-grade Python framework (agents, tools, handoffs, guardrails, tracing, output_type). (4) "Mastra Prompting Guide" — TypeScript-first framework (agents, workflows, tools, memory, RAG, evals; Vercel AI SDK base). Each post includes comparison tables vs the others. ~12,250 words combined
  • 🚀Enhancement10 new glossary entries supporting the agent-frameworks wave: langgraph, crewai, openai-agents-sdk, mastra, agent-graph, agent-handoff, agent-orchestration, agent-tool-loop, multi-agent-system, swarm. Glossary now at 195 entries
v2.47.0
  • ✨Feature4 new tutorials targeting the AI coding agents trending cluster — gap-fillers around the existing pillar. (1) "Vibe Coding: The Complete Guide (2026)" — definitive guide on the Karpathy-popularized term, when it works, when it breaks, prompt patterns, eval discipline, and the spec-driven alternative. (2) "Cline Prompting Guide" — fills the missing tool slot for the open-source VS Code coding agent (plan/act mode, MCP integration, multi-provider model config, cost/safety controls). (3) "Test-Driven Development With AI Coding Agents" — workflow tutorial for the red/green/refactor loop with Cursor/Claude Code/Cline, with prompt skeletons per phase. (4) "AI Coding Agent Evals: SWE-Bench, Aider Polyglot, Terminal-Bench" — reference tutorial on the three major public benchmarks and how to build internal evals that actually predict your team's experience. ~12,900 words combined
  • 🚀Enhancement10 new glossary entries supporting the AI coding agents wave: vibe-coding, agentic-coding, coding-agent, ai-ide, spec-driven-development, swe-bench, aider-polyglot, terminal-bench, codeact, cline. Glossary now at 185 entries
v2.46.0
  • ⚡ImprovementDirect-answer lead retrofits on 30 high-evergreen blog posts — strengthened or added the frontmatter `summary` field (renders as TL;DR card) on model how-to guides, head-to-head comparisons, foundational technique guides, framework playbooks, and advanced topic guides. Each summary is 40-80 words, declarative, AI-Overview-citable, with concrete tradeoffs/techniques named
  • ✨FeatureGlossary auto-linking infrastructure — the markdown parser now automatically wraps glossary terms (175 terms from lib/seo/glossary-data.ts) with anchor links to /glossary/<slug> at render time. Skips code blocks, existing links, headings; first-occurrence per page; longest-match-wins; case-insensitive; suppresses self-links on the term's own glossary page. Compiled-once regex, no per-render allocation. 20 unit tests covering all skip-zone and matching behaviors
v2.45.0
  • ✨Feature5 new canonical-tier definitive guides — completes Phase 3 net-new canonical scope. (1) "Prompt Evaluation: The Complete 2026 Guide" — eval pyramid, golden sets, LLM-as-judge, regression, observability. (2) "Fine-tuning vs Prompting vs RAG: The Complete 2026 Decision Guide" — three customization levers, decision framework, hybrid patterns. (3) "LLM Temperature and Sampling: The Complete 2026 Reference Guide" — temperature, top-p, top-k, penalties, seed, per-task recommendations, per-provider defaults table. (4) "Model Context Protocol (MCP): The Complete 2026 Guide" — protocol, primitives, security model, 2026 client/server ecosystem. (5) "Prompt Injection Defense: The Complete 2026 Security Guide" — direct, indirect, multi-modal injection variants with defense-in-depth mitigations. ~30,000 words combined
v2.44.0
  • ✨FeatureNew canonical pillar: "AI Voice and Audio Prompting — The Complete 2026 Guide." A 6,000-word entry point covering voice OUTPUT (TTS, voice cloning, music) and conversational voice (realtime speech-to-speech, voice agents) — the 2026 model landscape (ElevenLabs, OpenAI TTS, Realtime API, Hume, Cartesia, PlayHT, Google, open-weights), the universal voice-prompt anatomy, three architectures (one-shot TTS, voice cloning, realtime), voice-agent patterns, ethics/consent, and honest evaluation. Sister pillar to multimodal-input (which covers audio understanding). Completes Phase 3 modality coverage (image, video, reasoning, multimodal, enterprise, voice)
v2.43.0
  • ✨Feature3 new voice/audio tutorials: "Voice Generation Models Compared 2026" (ElevenLabs, OpenAI TTS, Hume, Cartesia, PlayHT, Google, open-weights), "Prompting GPT-4o Realtime Voice — A Speech-to-Speech Walkthrough" (session config, turn detection, tool calling, worked support agent), and "Audio Understanding with Gemini Long-Context — A Walkthrough" (long-form audio input with 5 shippable patterns)
  • ✨Feature6 new voice/audio glossary entries: text-to-speech (TTS), speech-to-text (STT) / automatic speech recognition, voice cloning, speaker diarization, prosody, and realtime voice API. Glossary now at 175 entries
v2.42.0
  • ✨FeatureNew canonical pillar: "Enterprise AI Adoption — The Complete 2026 Operating Model Guide." A 6,500-word entry point at the org-level operating-model layer (governance, ROI, budgeting, fluency programs, security, compliance, vendor strategy, build-vs-buy, SMB-to-enterprise adoption arc) — sister pillar to the existing function-patterns pillar (which covers what individual functions should prompt). Consolidates 9 enterprise/team adoption posts (budgeting, fluency, security, compliance, teams, SMB, business, local-business, ethics)
v2.41.0
  • ⚡ImprovementConversational prompt refinement is now a Pro-only feature. Free and anonymous users see an upsell on the AI Prompt Generator and can upgrade to unlock 3 follow-up refinements per prompt. Pro users keep their existing 3 refinements per prompt with a 100/day cap
  • ⚡ImprovementFree-tier AI Prompt Generator limit tightened from 5 to 3 generations per month. Anonymous (1/month) and Pro (200/month) limits are unchanged
v2.40.0
  • ✨FeatureNew canonical pillar: "Multimodal AI Prompting — The Complete 2026 Input Guide." A 5,400-word entry point for the multimodal INPUT side of prompting (sending images, PDFs, audio, video into text models) — covers the 2026 model landscape (GPT-4o, Claude Opus 4.7, Gemini 2.5 Pro/Flash), the universal multimodal-prompt anatomy, per-modality dialects (images, PDFs, audio, video, charts), shippable workflows (screenshot-to-code, document-to-data, photo-to-listing, whiteboard-to-spec, video-to-summary), and honest evaluation. Sister pillar to the image and video output pillars
v2.39.0
  • ✨FeatureNew canonical pillar: "AI Reasoning Models — The Complete 2026 Prompting Guide." A 6,800-word consolidated entry point covering what reasoning models actually are, the 2026 model landscape (o3, o4-mini, Claude Opus 4.7 and Sonnet 4.6 extended thinking, Gemini 2.5 Pro/Flash Deep Think, DeepSeek R1, Qwen QwQ), the universal reasoning-prompt anatomy, per-model dialects, why 2023-era chain-of-thought prompting backfires, thinking-budget control, hybrid workflows, and honest evaluation
v2.38.0
  • ✨FeatureNew canonical pillar: "AI Video Prompting — The Complete 2026 Guide." A 6,100-word consolidated entry point covering the extended video-prompt anatomy (image six-slot plus motion, camera, duration, audio), the 2026 model landscape (Veo 3, Sora 2, Runway Gen-3, Kling, Luma, Pika, Hailuo), per-model dialects, camera and motion vocabulary, multi-shot sequencing, audio prompting, and honest evaluation
v2.37.0
  • ✨FeatureNew canonical pillar: "AI Image Prompting — The Complete 2026 Guide." A 5,500-word consolidated entry point covering the universal six-slot prompt anatomy, the 2026 model landscape (Midjourney V7, DALL-E, Flux Pro, Stable Diffusion, Imagen, Ideogram, Firefly), per-model dialects, advanced control (character consistency, seeds, negative prompts), niche workflows, and honest evaluation
v2.36.0
  • ✨Feature4 new tutorials: agentic RAG walkthrough, corrective RAG (CRAG) implementation guide, HyDE retrieval walkthrough, and a semantic router implementation guide
  • ✨Feature10 more retrieval internals and long-context eval glossary entries: BM25, reciprocal rank fusion, contextual retrieval, cross-encoder, bi-encoder, ColBERT, many-shot jailbreaking, multi-query retrieval, needle in a haystack, and lost in the middle
v2.35.0
  • ✨Feature4 new RAG technique tutorials: chunking strategies for RAG, reranking retrieval walkthrough, hybrid search implementation guide, and a GraphRAG introduction guide
  • ✨Feature10 more RAG and agent glossary entries: agentic RAG, corrective RAG (CRAG), self-RAG, contextual compression, parent-document retrieval, semantic router, multimodal RAG, long-term memory, context rot, and document AI
v2.34.0
  • ✨Feature4 new technique tutorials: Least-to-Most prompting worked example, Step-Back prompting worked example, Chain-of-Code walkthrough, and a RAGAS evaluation walkthrough
  • ✨Feature10 more RAG-adjacent glossary entries: chunking, reranking, HyDE, embedding model, hybrid search, query rewriting, RLAIF, mixture of prompts, self-critique, and GraphRAG
v2.33.0
  • ✨Feature4 new technique tutorials: Chain-of-Density worked example, Program-of-Thoughts worked example, Self-Ask prompting guide, and a DSPy introduction guide
  • ✨Feature10 more glossary entries: least-to-most prompting, auto-CoT, step-back prompting, active prompting, chain of code, self-debug, model cascade, prefix-tuning, RAGAS, and structured decoding
v2.32.0
  • ✨Feature4 new applied-framework tutorials: scoring a customer service prompt with the SurePrompts Quality Rubric, 10 RCAF prompt templates for business tasks, a Context Engineering Maturity Model 30-minute workshop, and an Agentic Prompt Stack research-agent walkthrough
  • ✨Feature10 more glossary entries: chain of density, program of thoughts, self-ask, ReWOO, RAFT, eval harness, golden set, indirect prompt injection, DSPy, and prompt observability
v2.31.0
  • ✨FeatureNew Claude Opus 4.7 prompting guide covering extended thinking, 1M context, prompt caching, and agentic patterns
  • ✨FeatureNew LLM-as-Judge prompting guide — practical patterns for automating prompt evaluation, paired with the SurePrompts Quality Rubric
  • ✨Feature10 new glossary entries covering 2026 vocabulary: extended thinking, computer use, voice prompting, semantic caching, LLM-as-judge, plan-and-execute, self-refine, reflexion, skeleton of thought, and tool choice
v2.30.0
  • ⚡ImprovementUpdated AI Prompt Generator monthly limits: 1 generation for visitors, 5 for signed-in free accounts, 200 for Pro
v2.29.0
  • ✨FeaturePublished 4 named prompt engineering frameworks — SurePrompts Quality Rubric, RCAF Prompt Structure, Context Engineering Maturity Model, and Agentic Prompt Stack
  • ✨FeatureAdded 4 matching glossary entries for the new frameworks to complement the canonical blog posts
  • ⚡ImprovementAdded explicit AI crawler allowlist (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, and 5 others) to robots.txt
  • ⚡ImprovementExpanded llms.txt and added llms-full.txt for full-text LLM citation surface
  • ⚡ImprovementAdded version history to canonical-tier blog pages
v2.28.0
  • ⚡ImprovementBilling overhaul — adopted the official @freemius/sdk + React Starter Kit: in-app customer portal now supports cancel, upgrade, invoice download, and billing edit without leaving SurePrompts
  • ⚡ImprovementNew entitlement-centric schema (user_fs_entitlement) backed by the SDK’s purchase and webhook services for faster, more reliable Pro access sync
  • ✨FeatureSDK-powered webhook listener now handles license.extended, license.shortened, license.plan.changed, and license.deleted events in addition to the existing subscription events
v2.27.0
  • ✨FeatureSEO content expansion foundation — 3 new pillar articles on AI coding agents, context engineering, and prompt engineering for business teams
  • ✨Feature2 new glossary entries (context engineering, ReAct prompting) plus enhanced cross-linking for agentic AI, tool use, and prompt caching
  • ✨Feature20 new blog posts on prompting AI coding agents — Claude Code, Cursor, Devin, Replit Agent, Copilot Workspace, v0, Bolt.new, Windsurf, Aider, Continue.dev, ReAct, plan-and-execute, multi-agent, tool-use patterns, self-refine, reflexion, spec-driven coding, agent debugging, autonomous testing, and review agents
  • ✨Feature18 new blog posts on context engineering — prompt caching, long context, memory systems, token economics, retrieval patterns, hierarchical loading, context rot, needle-in-a-haystack, and the complete best-practices checklist
  • ✨Feature12 new blog posts on prompt engineering for business teams — marketing briefs, competitor analysis, campaign copy, sales discovery and proposals, pipeline forecasting, architecture reviews, incident postmortems, technical specs, SOPs, vendor evaluation, and process automation
v2.26.0
  • 🐛FixFixed pricing page showing "0+ templates" instead of actual template counts
  • 🐛FixFixed support widget crash when clicking the floating help icon
  • ✨Feature41 new blog posts — model guides, prompt patterns, industry deep dives (legal, healthcare, finance), product tutorials, and cost optimization
  • ✨Feature3 new competitor comparison pages
  • ✨FeatureAEO/GEO optimizations — TL;DR summaries, FAQ schema, SpeakableSchema, and ai.txt for AI answer engine visibility
  • ⚡ImprovementSEO improvements — canonical URLs, H1 hierarchy, hreflang, RSS feed fixes, and internal cross-linking
  • ⚡ImprovementAccessibility improvements — aria-labels on icon buttons and improved image alt text
  • 🚀Enhancement65 new tests for subscription, webhook, and admin endpoints
v2.25.0
  • 🐛FixRemoved fabricated structured data ratings and unverified social proof claims for content integrity
  • 🐛FixFixed race condition in prompt processing that caused double-rendering on template load
  • 🐛FixFixed prompt save confirmation showing success even when save failed
  • 🐛FixFixed rate limiter ignoring per-endpoint config when Upstash Redis is configured
  • 🐛FixFixed CSRF token validation not working in development mode
  • ⚡ImprovementAdded CSRF protection to subscription activation, prompt sharing, and community template publishing
  • ⚡ImprovementProduction error and warning logs are now always visible instead of being silenced
  • ⚡ImprovementEliminated theme flash — page no longer hidden until JavaScript loads
  • ⚡ImprovementOptimized database queries for user stats, tags, and prompt storage to reduce load times
  • 🐛FixFixed quality score of 0 being incorrectly stored as null
  • 🐛FixFixed encrypted storage data being inaccessible across browser tabs
  • ⚡ImprovementAdded rate limiting to community templates, shared prompts, and prompt libraries endpoints
  • ⚡ImprovementAdded search debouncing to prompt library model pages
  • 🐛FixQualified unsourced statistics across blog posts for content accuracy
v2.24.0
  • ✨Feature4 new Midjourney V7 blog posts — V7 vs Sora 2 vs Runway Gen-3 vs Veo 3 video AI comparison, plus persona guides for product photographers, fashion editorial, and animation/VFX workflows
  • ✨Feature40 new curated Midjourney V7 prompts seeded into /prompt-libraries/midjourney-v7 — 20 image prompts and 20 video prompts with camera-movement vocabulary and verified V7 parameters
  • ✨Feature6 new V7-specific templates in the template library — product shot, cinematic still, character portrait, video hero, video product, and video atmosphere, each with real V7 parameter controls
  • ⚡ImprovementRefreshed Midjourney prompt generator and builder landing copy to lead with V7 video capability (up to 21 seconds, camera-movement vocabulary, V7 parameter controls)
v2.23.0
  • ✨Feature4 new Grok blog posts — a real-time intelligence pillar guide plus persona guides for journalists, traders, and marketers, covering live X data workflows, DeepSearch research, and sentiment analysis
  • ✨Feature4 new Grok persona landing pages — /grok-prompt-generator-for-social-media, -journalists, -marketers, and -traders
  • ✨FeatureGrok added as a first-class prompt library model at /prompt-libraries/grok, with 30 curated prompts across real-time intelligence, trend analysis, social sentiment, news research, and creative categories
  • ✨Feature2 new glossary entries — Grok and xAI — with full definitions, related terms, and internal linking
v2.22.0
  • ✨FeatureNew deep-dive blog post — Advanced Prompt Engineering in 2026 covering Claude 4.6 adaptive thinking, GPT-5.4 reasoning effort levels, and Gemini 2.5 Deep Think, with a list of 2023-era techniques that stopped working on reasoning models
v2.21.0
  • ⚡ImprovementOne-click unsubscribe added to all SurePrompts emails — RFC 8058 List-Unsubscribe headers, dedicated unsubscribe page with click-to-confirm, and CAN-SPAM compliance for improved Gmail/Yahoo deliverability
v2.20.0
  • ✨Feature38 new glossary terms — embeddings, transformers, RLHF, jailbreaking, red-teaming, model distillation, vector databases, prompt caching, JSON mode, vision-language models, and more (70 total)
v2.19.0
  • ✨Feature10 new blog posts — Prompt Engineering Career Guide, Best AI Prompt Libraries 2026, AI Prompts for Personal Development/Construction/Local Business/Self-Reflection/Real Estate Investors/Supply Chain, How to Build a Prompt Library, and Prompt Engineering Certifications Guide
  • ✨Feature5 new competitor comparison pages — SurePrompts vs ChatGPT, Claude, Notion AI, Grammarly, and Rytr
  • ✨Feature3 new generator pages — Stable Diffusion, DALL-E, and ChatGPT Prompt Generators
  • ✨Feature2 new persona pages — AI Prompts for Coaches and Construction Professionals
  • ⚡ImprovementInternal linking round 4 — 20 new cross-links from 14 existing posts to all 10 new posts
v2.18.0
  • ✨Feature18 new blog posts including Best ChatGPT/Claude/Gemini/Grok/DeepSeek Prompts 2026, AI Prompts for SEO/Data Analysis/E-Commerce/Presentations/Job Interviews/Nonprofits/UX Designers, Prompt Chaining Guide, AI Image Prompt Guide, AI Prompt Formulas, Custom GPTs Guide, and Zero-Shot vs Few-Shot Prompting
  • ✨Feature12 new landing pages — 9 use-case generators (Ad Copy, SEO, Story, Data Analysis, Video Script, Newsletter, Landing Page, Image, Customer Service) plus Free AI Prompt Generator, Prompt Optimizer, and Midjourney Prompt Generator
  • ✨Feature7 new persona pages — AI Prompts for Accountants, Product Managers, Lawyers, E-Commerce Sellers, Data Analysts, Content Creators, and Customer Service teams
  • ✨Feature7 new competitor comparison pages — SurePrompts vs Jasper, Writesonic, Originality.ai, PromptBuilder, DocsBot, God of Prompt, and QuillBot
  • ⚡ImprovementInternal linking — 50+ cross-links across blog posts connecting generators, personas, alternatives, and related content
  • ⚡ImprovementConversion CTA section added to all 17 prompts-for persona pages
  • ⚡ImprovementNew test coverage for all SEO data files — 37 tests validating slug format, meta descriptions, cross-references, and data integrity
v2.16.0
  • ✨FeatureTemplate directory — browse all 320+ prompt templates organized by category at /templates, with per-category pages showing Free/Pro badges
  • ✨Feature6 new persona landing pages — AI Prompt Generator for Product Managers, Designers, Lawyers, Sales Professionals, Data Scientists, and Content Creators
  • ✨Feature10 new comparison blog posts — Claude vs Gemini, GPT-4o vs Claude Sonnet, Midjourney vs DALL-E, Copilot vs ChatGPT, Perplexity vs ChatGPT, DeepSeek vs ChatGPT, Grok vs ChatGPT, Claude vs ChatGPT for Coding, Gemini vs ChatGPT, and Llama vs ChatGPT
  • ⚡ImprovementSEO improvements — canonical URLs on 14 pages, OG/Twitter cards on key pages, SoftwareApplication structured data on builder pages, internal linking across all 72 blog posts
  • 🐛FixFixed duplicate H1 tags on 16 blog posts flagged by Bing Webmaster Tools
  • 🐛FixFixed short meta descriptions on 5 guide pages flagged by Bing
  • 🐛FixFixed missing H1 tags on /builder, /veo3-builder, and /template-builder
  • 🐛FixUnblocked /template-builder from robots.txt — now a proper SEO page with full metadata
v2.15.0
  • ✨FeaturePrompt refinement — after generating a prompt, refine it with follow-up instructions like "make it shorter" or "add more detail". 3 free refinements per generation.
  • ✨Feature8 new persona landing pages — AI Prompt Generator for Marketers, Students, Developers, Writers, Teachers, Business, HR, and Researchers
  • ✨Feature5 new blog posts including "How to Write AI Prompts", "How to Use Claude Like a Pro", and "10 Best AI Prompt Frameworks"
  • ⚡ImprovementUpdated all dependencies to latest versions including React 19.2, Supabase 2.100, and Zod 4.3
  • 🐛FixSecurity patches for Next.js (HTTP request smuggling, image cache exhaustion) and flatted (prototype pollution)
v2.14.0
  • ✨FeatureSupport system — signed-in users can submit support requests via a dedicated /support page
  • ✨FeatureFloating support widget — quick-access support form available from any page for authenticated users
  • ✨FeatureAdmin support dashboard — view, filter, and manage support tickets with status tracking and internal notes
v2.13.0
  • ✨FeatureThree-tier AI generation limits — anonymous users get 2/month, signed-in free users get 10/month, Pro users get 200/month
  • ✨FeatureServer-side generation limit enforcement — limits are now tracked on the server and cannot be bypassed by clearing browser data
  • ✨FeatureSmart upgrade modals — anonymous users see a "Sign up free for 10/month" prompt, signed-in users see the Pro upgrade offer
  • 🚀EnhancementNew signups now redirect to the AI Prompt Generator instead of the template builder
  • 🚀EnhancementNavigation label updated from "AI Generator" to "AI Prompt Generator"
  • ⚡ImprovementUsage indicator adapts to auth state — shows relevant upgrade path for each user tier
v2.12.0
  • ✨FeatureInline AI Prompt Generator — full textarea + streaming output embedded on the page, no modal needed
  • ✨FeatureSave AI-generated prompts to My Prompts with one click
  • ✨FeatureAI Generated filter in My Prompts to quickly find AI-created prompts
  • ✨FeatureRecent Generations section on AI Generator page shows last 5 saved prompts
  • ✨FeatureRegenerate with AI — re-open the generator with the original request from any saved AI prompt
  • ✨FeatureSession-based generation history — browse and re-copy recent generations without signing in
  • ✨FeatureTemplate suggestions after AI generation — discover relevant templates from the builder
  • ✨Feature3 new AI models: Perplexity, DeepSeek, and Copilot (total: 9 models)
  • ✨Feature9 new Prompt Builder landing pages for model-specific template building
  • 🚀EnhancementAI Prompt Generator is now a standalone top-level nav item
  • 🚀EnhancementGenerator landing pages link to AI Generator with model pre-selected
  • 🚀EnhancementBuilder landing pages include Builder vs Generator comparison section
  • 🚀EnhancementCross-link banners between generator and builder pages
  • 🚀EnhancementAI badge on prompt cards, detail modals, and shared prompt pages
  • 🚀EnhancementDynamic Open Graph images for AI Generator model pages
  • ⚡ImprovementPremium templates split into separate chunk — 45% smaller initial bundle
  • ⚡ImprovementFooter expanded with all 9 builder page links
  • ⚡ImprovementSitemap includes all builder and generator landing pages
v2.11.0
  • ✨FeatureAI Prompt Generator is now the primary product — all CTAs, navigation, and messaging lead with AI generation
  • 🚀EnhancementNavigation renamed from "Prompt Builders" to "Prompt Tools" with AI Generator first
  • 🚀EnhancementHomepage hero, How It Works, and final CTA all center on AI Prompt Generator
  • 🚀EnhancementTemplate Builder repositioned as complementary tool with 320+ templates
  • ⚡ImprovementUpdated all site-wide CTAs to direct users to AI Prompt Generator
  • ⚡ImprovementRefreshed metadata, SEO, and structured data to lead with AI generation
v2.10.0
  • ✨FeatureAI Prompt Generator promoted across the entire site — homepage spotlight, pricing integration, footer, and about page
  • ✨FeatureNew blog post: AI Prompt Generator guide with step-by-step walkthrough and real examples
  • ✨FeatureNew blog post: Template Builder vs. AI Prompt Generator comparison with decision framework
  • 🚀EnhancementPricing page now shows AI Prompt Generator in both Free and Pro plans with generation limits
  • 🚀EnhancementPricing page features new AI Generator spotlight banner between plans and comparison table
  • 🚀EnhancementHomepage hero links directly to AI Prompt Generator as an alternative to template builder
  • 🚀EnhancementHomepage final CTA now includes AI Generator alongside the main builder CTA
  • ⚡ImprovementFeature comparison table on pricing page includes AI Prompt Generator row (2/month free, 100/month pro)
v2.9.0
  • ⚡ImprovementFixed stale template counts across entire site — now computed from source data (320+ total, 110+ free, 210+ pro)
  • ✨FeatureConversion analytics — track pricing page views, upgrade modal impressions, checkout starts/completions
  • 🚀EnhancementRedesigned upgrade modal with live pricing, annual/monthly toggle, testimonial, and trust signals
  • 🚀EnhancementBuilder template selector now shows Pro upgrade nudge with hover overlay on locked templates
  • 🚀EnhancementPricing page now includes testimonials, FAQ section, and annual savings emphasis
  • ⚡ImprovementJSON builder upgrade modal consolidated to shared component
v2.8.0
  • ✨FeatureEnhanced prompt scoring with 4 sub-categories: Completeness, Specificity, Structure, Enhancement
  • ✨FeatureVisual strength meter with animated SVG circle and sub-score breakdown bars
  • ✨FeatureReal-time field-level hints — inline feedback when inputs are vague or too short
  • ✨FeatureAuto-save drafts — your work is saved locally and restored within 24 hours
  • ✨FeatureKeyboard shortcuts: ⌘+S save, ⌘+Shift+C copy, ⌘+Shift+R reset, ⌘+K search templates
  • ⚡ImprovementRecent templates tracked for quick access
v2.7.0
  • ✨Feature4 alternative comparison pages: SurePrompts vs PromptPerfect, PromptBase, PromptHero, and FlowGPT
  • ✨FeatureAlternatives index page with overview of all comparisons
  • ⚡ImprovementInternal linking from 26 blog posts to use-case prompt generators (~90 new links)
v2.6.0
  • ✨Feature8 use-case prompt generator landing pages: email, blog post, resume, social media, business plan, cover letter, code, and product description
  • ⚡ImprovementEach use-case page includes tips, templates, FAQs, and cross-links to related generators
v2.5.0
  • ✨Feature3 model comparison landing pages: ChatGPT vs Claude, ChatGPT vs Gemini, Claude vs Gemini
  • ⚡ImprovementModel generators added to header dropdown and footer navigation
  • ⚡ImprovementInternal linking from 25 blog posts to model-specific prompt generators
v2.4.0
  • ✨FeatureModel-specific landing pages: ChatGPT, Claude, Gemini, Midjourney, and Copilot prompt generators
  • ⚡ImprovementEach model page includes tailored tips, FAQs, template recommendations, and cross-links
  • ⚡ImprovementFAQSchema structured data on all model landing pages for rich search snippets
v2.3.0
  • ✨Feature5 new role-specific blog posts: AI prompts for lawyers, project managers, sales, finance, and designers
  • ✨Feature4 new product guides: JSON Prompt Builder, Veo 3 Builder, Prompt Optimizer, and Team Collaboration
  • ⚡ImprovementHowTo structured data added for all new guides
v2.2.0
  • ✨FeatureRich structured data: WebSite with sitelinks searchbox, Product schema on pricing, ItemList on template galleries, HowTo on guides, and AggregateRating for search result enhancements
  • ✨FeatureFAQ sections with schema markup added to about, glossary, optimizer, compare, and guides pages
  • ⚡ImprovementExpanded social proof section with 6 role-specific testimonials and trust signal bar
  • ⚡ImprovementPricing page now includes detailed feature comparison table and trust signals (money-back guarantee, secure payment, cancel anytime)
v2.1.0
  • ✨FeatureCustom Template Builder: create your own prompt templates with custom fields, live preview, and reusable skeleton templates
  • ✨FeatureCustom templates integrate with the main builder — use them just like built-in templates
  • ⚡ImprovementTemplate builder supports all field types: text, multiline, dropdown, multi-select, and chips
  • ⚡ImprovementQuick-insert placeholders in skeleton editor for fast template authoring
v2.0.0
  • ✨FeaturePrompt A/B Compare: side-by-side and diff view for comparing any two saved prompts
  • ✨FeatureUsage Analytics Dashboard: track prompt count, quality trends, template usage, and weekly activity
  • ✨FeatureHomepage Quick Access: recent prompts and favorites surfaced on the homepage for logged-in users
  • ⚡ImprovementDashboard link added to navigation for authenticated users
  • ⚡ImprovementQuick links to Dashboard, Compare, and Community from homepage
v1.9.0
  • ✨FeaturePrompt Sharing: share prompts via public link with view/copy tracking and OG metadata
  • ✨FeatureExport prompts to Markdown, plain text, or JSON with full metadata
  • ✨FeaturePrompt Version History: track iterations with diff view and restore to previous versions
  • ✨FeatureCommunity Templates: browse, search, upvote, and fork prompts shared by other users
  • ✨FeatureTeam Workspaces: create teams, invite members with role-based access, and share prompt libraries
  • ⚡ImprovementShare & Export modal with copy-to-clipboard for plain text and markdown formats
  • ⚡ImprovementCommunity template gallery with category filters, sorting, and pagination
v1.8.0
  • ✨FeatureProduction error monitoring: automatic client-side error capture with batched reporting, deduplication, and breadcrumb context
  • ✨Feature5 new free templates: Code Review, AI Persona Creator, Decision Framework, Interview Prep, and Learning Plan
  • ⚡ImprovementErrorBoundary now reports errors to monitoring service and uses consistent teal theme
  • ⚡ImprovementOptimized OG image from 2.7MB to 687KB and team photos for faster page loads
  • ⚡ImprovementAdded preconnect hints for fonts to speed up initial render
  • ⚡ImprovementExpanded test coverage with new tests for ErrorBoundary, CookieConsent, and error reporting
  • 🐛FixCSP violation reports now use structured logging for better server-side observability
v1.7.0
  • ✨FeaturePrompt Optimizer: paste any prompt to get instant quality analysis, scoring, and one-click improvements
  • ✨Feature8-category scoring engine analyzes role, specificity, structure, output format, constraints, context, and examples
  • ✨FeatureAuto-improve toggles: add role framing, structure, output format, and model-specific optimization
v1.6.3
  • ⚡ImprovementLazy-load Builder, JSON Builder, Veo 3 Builder, and My Prompts pages for faster initial page loads
  • ⚡ImprovementOptimized package imports for date-fns and zod to reduce bundle size
v1.6.2
  • ⚡ImprovementImproved accessibility with ARIA labels on interactive elements, keyboard navigation, and proper heading hierarchy
  • ⚡ImprovementReplaced production console logging with dev-only logger across all components and API routes
v1.6.1
  • ⚡ImprovementAdded branded 404 page with navigation to Builder and Templates
  • ⚡ImprovementAdded custom error page with retry functionality for unexpected errors
v1.6.0
  • ✨FeaturePrompt Variants: generate Concise, Detailed, and Expert versions of every prompt (Pro)
  • ✨FeatureModel-Optimized Output: prompts tuned for Claude, GPT-4, Gemini, and Llama (Pro)
  • ✨FeatureSmart Presets: one-click field pre-fills for popular templates (Pro)
  • ✨FeaturePrompt Chains: multi-step prompt workflows for complex tasks (Pro)
v1.5.0
  • ✨FeatureAdded subcategory tags to all 170+ templates for easier discovery
  • ✨FeatureAdded "Recommended for You" section based on your usage history
  • ✨FeatureAdded sorting options: Recommended, Newest, and A-Z
  • 🚀EnhancementTemplates now show Popular and New badges for quick identification
  • ⚡ImprovementSearch now matches against template tags for better results
v1.2.0
  • ✨FeatureAdded 14 new trending AI use case templates covering coding tools, MCP, multi-agent systems, AI safety, voice AI, and more
  • ✨FeatureAdded 3 free templates: AI Coding Assistant Rules, Prompt Chain Builder, and AI Voice Agent Script
  • ✨FeatureAdded 11 pro templates including MCP Server Designer, Multi-Agent Orchestrator, AI Feature PRD, LLM Evaluation Framework, and more
v1.1.0
  • ✨FeatureNew lower pricing: Pro plan now $3.99/month or $29.99/year
  • ✨FeatureAdded yearly billing option with 37% savings
  • ⚡ImprovementImproved checkout experience with streamlined billing
v1.0.7
  • ✨FeatureAdded 50 new comprehensive prompt templates across free and premium tiers
  • ✨FeatureAdded 15 new free templates: Social Media Campaign, Email Newsletter, Brand Voice Guide, Blog Content Outline, Article Writer, Content Calendar, Literature Review, Competitive Analysis, Trend Report, Meeting Notes, Process Documentation, Project Brief, Prompt Refinement, AI Use Case Explorer, and Automation Planner
  • ✨FeatureAdded 5 business strategy templates: Market Entry Strategy, Pivot Analysis, Partnership Evaluation, Business Model Canvas, and Competitive Positioning
  • ✨FeatureAdded 5 technical writing templates: API Tutorial Writer, Release Notes Generator, Troubleshooting Guide Creator, User Guide Documentation, and Knowledge Base Article Writer
  • ✨FeatureAdded 5 data analysis templates: Data Visualization Recommendations, A/B Test Analysis Report, Dashboard Design Specifications, and Statistical Analysis Report
  • ✨FeatureAdded 5 AI automation templates: Workflow Automation Designer, AI Agent Architecture, System Integration Planner, ML Model Evaluation Framework, and Chatbot Conversation Designer
  • ✨FeatureAdded 10 video production templates: Animated Explainer Video Script, Commercial/Ad Script, Social Media Short Video, Product Demo Video, Tutorial/Educational Video, Documentary Sequence, Event Highlight Reel, Brand Story Video, Music Video Concept, and Behind-the-Scenes Content
  • ✨FeatureAdded 5 advanced sales templates: Objection Handling Framework, Discovery Question Generator, Sales Proposal/RFP Writer, Demo Script Creator, and Negotiation Strategy Planner
  • 🚀EnhancementAll templates follow production-ready standards with comprehensive field configurations and Mustache syntax support
  • 🚀EnhancementPremium templates include detailed skeletons with industry best practices and strategic frameworks
  • ⚡ImprovementExpanded template library now offers over 140 total templates for diverse use cases
v1.4.2
  • ✨FeatureAdded 50 professional Midjourney V7 prompts following official Midjourney V7 parameter syntax and guidelines
  • ✨FeatureAdded 50 professional Flux Pro prompts for Black Forest Labs text-to-image generation
  • ✨FeatureMidjourney V7 prompts include cinematic (12), product (10), abstract (10), character (10), and nature (8) categories
  • ✨FeatureFlux Pro prompts include photorealistic portraits (12), product photography (10), architecture & interiors (10), fantasy & sci-fi (10), and nature photography (8)
  • 🚀EnhancementMidjourney V7 prompts use authentic V7 parameters: --ar, --s, --chaos, --seed, --v 7, --oref, --ow, --no
  • 🚀EnhancementFlux Pro prompts optimized for static image generation with technical photography details and API parameters
  • ⚡ImprovementAll prompts researched from official documentation and 2025 best practices for accurate model-specific formats
v1.4.1
  • ✨FeatureAdded 50 professional Runway Gen-3 Alpha prompts following official Runway guidelines across 5 categories
  • ✨FeatureIncluded cinematic (12), product demo (10), abstract (10), character (10), and nature (8) prompts
  • 🚀EnhancementPrompts follow Runway [camera]: [scene]. [details] structure with 100-250 word length
  • 🚀EnhancementCamera movements include FPV, tracking, orbital, push-in, crane, dolly zoom, whip pan, and steadicam shots
  • ⚡ImprovementTechnical parameters stored in JSONB: duration (5s/10s), aspect ratio (16:9/9:16/1:1), resolution, model type
v1.4.0
  • ✨FeatureAdded Veo 3 Video Prompt Builder with dual text/JSON output format
  • ✨FeatureIntroduced quick-start presets: Cinematic Product Shot, Documentary Style, and Abstract Art
  • ✨FeatureAdded real-time JSON generation matching Google Veo 3 API specifications
  • 🚀EnhancementNew Video AI category with specialized video generation templates
  • ✨FeatureSide-by-side text and JSON preview with independent copy buttons
  • 🚀EnhancementComprehensive form fields including camera movement, lighting, scene elements, and technical parameters
v1.3.0
  • ✨FeatureAdded 50 professional Sora 2 prompts following OpenAI guidelines across 5 categories
  • ✨FeatureIncluded abstract (12), cinematic (12), character (10), product demo (8), and nature (8) prompts
  • 🚀EnhancementOptimized prompts for Sora 2 API with shot-list format, detailed camera work, lighting, actions, and sound design
  • ⚡ImprovementPrompts include specific timing, color palettes, and professional cinematography terminology
v1.2.3
  • 🚀EnhancementRedesigned prompt library to use card-based grid layout for better visual scanning
  • ⚡ImprovementEnhanced card design with shadow elevation and hover effects for better engagement
  • 🚀EnhancementImproved responsive grid layout: 1 column mobile, 2 columns tablet, 3 columns desktop
v1.2.2
  • ✨FeatureRedesigned prompt library to use list→detail pattern with individual prompt pages
  • ✨FeatureAdded previous/next navigation for browsing prompts
  • ✨FeatureAdded related prompts section on detail pages
  • 🚀EnhancementImproved prompt library performance with lightweight list view
  • 🚀EnhancementAdded SEO-optimized individual URLs for each prompt
v1.2.1
  • ✨FeatureAdded hybrid text/JSON format display for Veo 3 prompts with format toggle
  • ✨FeatureImplemented structured JSON prompt format matching Veo 3 API specifications
  • 🚀EnhancementAll 50 Veo 3 prompts now include API-compatible JSON representation with camera, lighting, elements, and keywords
  • ⚡ImprovementCopy button now indicates format type (Text or JSON) being copied
  • 🚀EnhancementAnalytics tracking enhanced to include prompt format (text/json) in copy events
v1.2.0
  • ✨FeatureLaunched Model-Specific Prompt Libraries for Veo 3 and Sora 2 (Pro feature)
  • ✨FeatureAdded curated, copy-ready prompts for video AI models with categories and filtering
  • ✨FeatureIntegrated usage tracking and analytics for prompt library interactions
  • 🚀EnhancementPro subscription now includes unlimited access to all prompt libraries
v1.1.5
  • 🐛FixFixed critical database security vulnerability in active_subscriptions view that could expose user subscription data
v1.1.4
  • 🚀EnhancementUpdated brand color scheme to modern navy-teal-sage palette across all pages
  • ✨FeatureAdded 27 new premium templates across 4 categories: AI Automation (7), Business Development (7), Growth Marketing (8), HR Recruiting Enhanced (5)
  • ⚡ImprovementUpdated landing pages (prompt-generator, ai-prompt-builder) and about page with new visual identity
  • 🚀EnhancementApplied consistent color theming to builder components and UI elements
v1.1.5
  • ✨FeatureAdded admin panel for manual IndexNow URL submission and sitemap indexing
  • 🚀EnhancementCreated bulk submission script for indexing entire sitemap with IndexNow
  • ⚡ImprovementAdded test endpoint and npm scripts for IndexNow verification
v1.1.4
  • ✨FeatureIntegrated IndexNow for instant search engine indexing of new content
  • 🚀EnhancementAdded API endpoint for automated URL submission to search engines
  • ⚡ImprovementImproved SEO with faster content discovery through IndexNow protocol
v1.1.3
  • 🚀EnhancementJSON Builder interface now blurs after free trial expires with clear upgrade prompt
  • ⚡ImprovementEnhanced trial expiration messaging with visual indicators
  • ✨FeatureAdded inline upgrade overlay for expired JSON Builder trials
v1.1.2
  • ⚡ImprovementJSON Prompt Builder now offers 1 free use before requiring Pro subscription
  • ✨FeatureAdded usage tracking for JSON Builder with automatic reset after 30 days
  • 🚀EnhancementImplemented upgrade modal with clear Pro benefits for JSON Builder
v1.1.1
  • 🚀EnhancementJSON Prompt Builder is now a Pro-only feature with premium access control
  • ⚡ImprovementAdded Pro badge to JSON Builder in navigation for better visibility
  • ✨FeatureImplemented upgrade prompt for free users accessing JSON Builder
v1.1.0
  • ✨FeatureIntroduced "Open in AI" buttons for direct integration with Claude, ChatGPT
  • ⚡ImprovementOptimized button layouts for single-line display on mobile devices
  • 🐛FixRemoved PerformanceMonitor component that was causing infinite re-renders
  • 🚀EnhancementUpdated color scheme across modals and panels for consistent branding
v1.0.10
  • ✨FeatureAdded JSON Prompt Builder for creating structured prompts in JSON format
  • ✨FeatureIncluded preset templates for common JSON prompt use cases (blog posts, code reviews, social media)
  • ✨FeatureAdded JSON export/download functionality with syntax highlighting
  • 🚀EnhancementImplemented auto-save functionality for JSON builder drafts
v1.0.9
  • ✨FeatureAdded Pro Only Advanced Options in the builder with upgrade prompts
  • ✨FeatureCompletely redesigned homepage with modern animations and interactive elements
  • ✨FeatureIntroduced new color scheme and dark mode
  • 🐛FixResolved React unescaped entity errors ensuring stable builds
  • ⚡ImprovementRedesigned header navigation with user account dropdown for cleaner UI
v1.0.8
  • ✨FeatureAdded 20 new premium templates across 6 categories (Financial, Sales, Customer Success, Product Management, Education, Crisis Management)
  • 🚀EnhancementRedesigned library page with color-coded category themes for better visual organization
  • ✨FeatureAdded three comprehensive blog posts on prompt engineering best practices
  • ⚡ImprovementEnhanced Google Analytics integration with privacy-first cookie consent
  • ⚡ImprovementOptimized app performance for faster template loading and smoother interactions
  • 🐛FixResolved display issues on the library page for better mobile experience
v1.0.7
  • 🚀EnhancementImproved authentication flow with auto-checkout support for logged-in users
  • 🐛FixFixed OAuth redirect flow to preserve intended destination after login
  • ⚡ImprovementEnhanced user experience from pricing page to checkout process
v1.0.6
  • ✨FeatureLaunched Pro and Premium subscription tiers with Stripe integration
  • ✨FeatureAdded 45+ premium templates across 10 categories for Pro subscribers
  • ✨FeatureIntroduced billing management and subscription portal
  • 🚀EnhancementImplemented comprehensive security enhancements including CSRF protection
  • ✨FeatureAdded automated testing suite with 60%+ code coverage
v1.0.5
  • ✨FeatureAdded "What's New" changelog panel to track app updates
  • ⚡ImprovementMade blog post content fully responsive on mobile devices
  • ⚡ImprovementLimited table of contents to H2 headings for better readability
  • 🐛FixFixed table of contents navigation links to match heading IDs
v1.0.4
  • ✨FeatureIntroduced blog system with markdown support
  • ✨FeatureAdded table of contents for blog posts
  • 🚀EnhancementImplemented syntax highlighting for code blocks
v1.0.3
  • ✨FeatureLaunched comprehensive guides section for prompt engineering
  • ⚡ImprovementEnhanced prompt quality scoring algorithm
  • 🐛FixResolved authentication issues with email verification
v1.0.2
  • ✨FeatureAdded admin dashboard for user management
  • 🚀EnhancementImproved template categorization and search
  • ⚡ImprovementOptimized database queries for faster load times
v1.0.1
  • ✨FeatureIntroduced dark mode with system preference detection
  • ⚡ImprovementEnhanced mobile responsiveness across all pages
  • 🐛FixFixed template rendering issues with special characters
v1.0.0
  • ✨FeatureLaunched SurePrompts - AI prompt builder platform
  • ✨FeatureAdded 20+ professional prompt templates
  • ✨FeatureImplemented real-time prompt enhancement engine
  • ✨FeatureCreated user authentication with Supabase
  • ✨FeatureAdded local and cloud storage for saved prompts