Skip to main content

DeepSeek V4

A large open-weight mixture-of-experts model with a 1M-token context and strong reasoning at a low price.

DeepSeek V4 specifications

ProviderDeepSeek
Context window1,000,000 tokens (1.0M)
Input price$0.43 / 1M tokens
Output price$0.87 / 1M tokens
InputsText
Reasoning modeYes
Vision (image input)No
Tool / function callingYes
Native real-time webNo
Open-weightYes

Pricing & context web-verified June 2026 confirm on DeepSeek's pricing page. V4 Pro tier; context-cache hits bill far less. A cheaper V4 Flash tier ($0.14 / $0.28 per 1M) is also available.

What DeepSeek V4 is best for

  • Cost-sensitive reasoning and coding
  • Long-context analysis at low cost
  • Self-hosted / open-weight deployments

How to prompt DeepSeek V4

  • Text-only — use a multimodal model if you need image input.
  • It reasons while calling tools — give it tools for agentic tasks.
  • Cache context to cut cost further; a cheaper V4 Flash tier exists for high-volume work.

Work with DeepSeek V4

Frequently asked questions

What is DeepSeek V4's context window?

DeepSeek V4 has a 1,000,000-token context window (about 1.0M tokens) — the maximum amount of text it can consider at once.

How much does DeepSeek V4 cost?

Via the DeepSeek API, DeepSeek V4 costs $0.43 per 1M input tokens and $0.87 per 1M output tokens (web-verified June 2026). Output tokens cost more, so the length of the model's reply usually drives the bill.

Is DeepSeek V4 multimodal?

No — DeepSeek V4 is text-only. Use a vision-capable model if you need to send images.

Is DeepSeek V4 open-weight?

Yes — DeepSeek V4 is open-weight, so you can self-host it or run it through a third-party host.