← All providers

DeepSeek API Pricing, Models & Rate Limits (2026)

DeepSeek trains its own models and is the aggressive-pricing end of the frontier field — V4 Pro ships a visible chain-of-thought thinking mode at a fraction of big-lab flagship pricing, and V4 Flash undercuts nearly every non-reasoning model on this site.

Current models
2
Legacy models
0
Price range /M
$0.17–$0.54
Max context
128K
Median tok/s
100
Next retirement
—

Models and prices

ModelInput /MOutput /MBlended /M
DeepSeek V4 Flash$0.14$0.28$0.17
DeepSeek V4 Pro$0.43$0.87$0.54

Speed

Fastest measured DeepSeek model is DeepSeek V4 Flash at 132 tokens/sec (280ms TTFT), median across measured DeepSeek models is 100 tokens/sec. See the full speed benchmark methodology.

Operational facts

Verified 2026-08-08 against source.

OpenAI-compatibleYes
API base URLhttps://api.deepseek.com/v1
Auth modelBearer API key
Prompt cachingYes
Batch discountNot documented
Free tierNone
Rate-limit modelNo published hard request caps; dynamic throttling under heavy load
Data residencyNot documented
Trains on API dataNot documented
SLA publishedNo
DocsOfficial pricingStatus page

Switching to and from DeepSeek

The closest-priced current model to DeepSeek V4 Pro ($0.54/M) outside DeepSeek is Gemini 3.5 Flash Lite ($0.56/M, +3.4%). Same request shape through All AI Ask — swap the model id, keep the rest of your code.
The closest-priced current model to DeepSeek V4 Flash ($0.17/M) outside DeepSeek is Ministral 8B ($0.15/M, -14.3%). Same request shape through All AI Ask — swap the model id, keep the rest of your code.

Calling DeepSeek through All AI Ask

curl https://allaiask.com/api/v1/chat \
  -H "Authorization: Bearer $ALLAIASK_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "deepseek-v4-flash", "messages": [{"role": "user", "content": "Hello"}]}'

FAQ

Is DeepSeek OpenAI-compatible?

Yes — DeepSeek's API base (https://api.deepseek.com/v1) accepts the OpenAI SDK request/response shape, so existing OpenAI client code works with a base-URL and key swap.

Does DeepSeek support prompt caching?

Yes, as of 2026-08-08 — see https://api-docs.deepseek.com for the current mechanics and discount.

Does DeepSeek have a free tier?

No published free tier as of 2026-08-08 — billing starts from the first request.

How much does the DeepSeek API cost?

Current DeepSeek models range from $0.17 to $0.54 per million blended tokens (3:1 input:output). Full per-model pricing is in the table below.

Where is DeepSeek API data hosted?

Not documented as of 2026-08-08 — no published data-residency commitment found for DeepSeek.

Try DeepSeek for free

Run real prompts against every current DeepSeek model, and every other provider on this site, in one workspace.

Try It Free