AI Model Watch
Providers / DeepSeek

DeepSeek AI models

Open-weight reasoning and chat models known for aggressive pricing.

Cheapest GADeepSeek-V4.1-Flash$0.525 /M blendedBiggest contextDeepSeek-V4.1-Flash1M tokensFlagshipDeepSeek-V4.1-FlashGATracked6 models2 generally available
Lifecycle watch · 4
The lineup

All DeepSeek models

Official pricing ↗
Model ↕Status ↕Context ↕Input $/M ↕Output $/M ↕Blended $/M ↕Cutoff ↕
DeepSeek-V4.1-Flashopen
DeepSeek · text, image
GA1M$0.3$1.20$0.525—
DeepSeek-V4-Proopen
DeepSeek · text
GA1M$1.32$3.96$1.98—
DeepSeek-V4-Flash-Vision-Expopen
DeepSeek · text, image
Retired1M$0.44$1.32$0.66—
DeepSeek-V4-Flashopen
DeepSeek · text
Retired1M$0.44$1.32$0.66—
deepseek-chat (legacy alias)open
DeepSeek · text
Retired1M$0.14$0.28$0.175—
deepseek-reasoner (legacy alias)open
DeepSeek · text
Retired1M$0.14$0.28$0.175—

Blended = 0.75 × input + 0.25 × output $/M tokens (a fair single-number cost proxy). Click any header to sort.

FAQ

DeepSeek pricing & models

What is the cheapest DeepSeek model?

DeepSeek-V4.1-Flash is the cheapest generally-available DeepSeek model we track, at $0.3 per 1M input tokens and $1.20 per 1M output tokens ($0.525/1M blended).

What is DeepSeek's flagship model?

DeepSeek-V4.1-Flash is DeepSeek's most prominent model in our catalog, with a 1M-token context window and pricing of $0.3/$1.20 per 1M input/output tokens.

How many DeepSeek models are there?

We track 6 DeepSeek models, of which 2 are generally available and 4 are deprecated or scheduled for retirement.

Which DeepSeek models are being deprecated?

deepseek-chat (legacy alias) (retires 24 Jul 2026), deepseek-reasoner (legacy alias) (retires 24 Jul 2026), DeepSeek-V4-Flash-Vision-Exp (retires 10 Sep 2026), DeepSeek-V4-Flash (retires 10 Sep 2026).