StackLift
DeepSeek · updated Nov 22, 2026, 4:07 PM UTC

DeepSeek V3 API pricing.

DeepSeek V3 by DeepSeek costs $0.257 per million input tokens and $1.03 per million output tokens, with a 164K-token context window. That puts it in the budget range of the models we track. A support chatbot handling 10,000 conversations a month would cost about $9.00 a month in API fees.

Price

What DeepSeek V3 costs per token

List prices, checked every hour. Blended assumes three input tokens for every output token. This model sits in the budget range of the models we track.

ModelProviderInput / 1MOutput / 1MBlended / 1MContext
DeepSeek V3DeepSeek$0.257$1.03$0.45164K

Source: DeepSeek V3 in the OpenRouter model catalog. Check it against DeepSeek's API pricing.

Specs

What DeepSeek V3 can do

From DeepSeek V3's entry in the OpenRouter model catalog.

Reads
text
Writes
text
Context window
164K tokens
Longest answer
16K tokens
Tool calling
Yes: it can call functions and APIs you define
Structured output
Yes: answers can follow a JSON schema
Reasoning
Not listed
Open weights
Yes: you can also run it on your own servers
Knowledge cutoff
July 2024

Monthly cost

What common workloads cost on DeepSeek V3

API fees only, at list price. The assumptions are shown, so you can scale them to your own volume.

Customer support chatbot

$9.00 / month

10,000 conversations a month, ~1,500 in and ~500 out each: 15M input and 5M output tokens.

Knowledge-base Q&A (RAG)

$28.82 / month

20,000 questions a month, ~4,000 in and ~400 out each: 80M input and 8M output tokens.

Document summaries and extraction

$7.82 / month

2,000 documents a month, ~12,000 in and ~800 out each: 24M input and 1.6M output tokens.

Coding assistant for a team of 10

$13.89 / month

10 developers a month, 3.0M in and 0.6M out each: 30M input and 6M output tokens.

Hosting, a vector database, monitoring and the build itself are extra. The free StackLift estimate adds them for your exact scope.

Per request

What one request costs on DeepSeek V3

A request costs its input tokens times the input price plus its output tokens times the output price. A typical assistant turn of 2,000 input and 500 output tokens on DeepSeek V3: 2,000 × $0.257 ÷ 1M + 500 × $1.03 ÷ 1M = $0.0010.

RequestInput / output tokensOne request1,000 requests
Short chat reply500 / 200$0.0003$0.334
Typical assistant turn2,000 / 500$0.0010$1.03
Long document task20,000 / 1,500$0.0067$6.69

As you grow

DeepSeek V3 cost at ten times less and ten times more

The same workloads at a tenth and at ten times the volume, at DeepSeek V3's list price.

WorkloadA tenthTypicalTen times
Customer support chatbot$0.901,000 conversations$9.0010,000 conversations$90.04100,000 conversations
Knowledge-base Q&A (RAG)$2.882,000 questions$28.8220,000 questions$288.22200,000 questions
Document summaries and extraction$0.78200 documents$7.822,000 documents$78.2420,000 documents

Fit

When DeepSeek V3 is the right choice

DeepSeek V3 is priced for volume: classification, extraction, routing and short replies, where thousands of requests a day must stay cheap. Hard reasoning may need a larger model for some steps.

DeepSeek V3 works with text and can call tools and APIs and return structured JSON, so it can drive agents that act in your systems.

Its weights are published, so DeepSeek V3 can also run on your own servers when data must stay in-house; the API price above is for hosted use.

Many products mix models: a cheaper one for simple steps and a stronger one where quality matters. Test a shortlist on your own examples before you commit.

Save

Ways to pay less for DeepSeek V3

  1. 1. Cache repeated instructions

    Most providers discount input that repeats across requests, such as a long system prompt or shared context.

  2. 2. Route by difficulty

    DeepSeek V4 Flash 0423 costs 75% less per token than DeepSeek V3. Send simple requests there and keep DeepSeek V3 for the steps that need it.

  3. 3. Batch what can wait

    Overnight or bulk jobs often qualify for batch pricing, at a lower rate than live requests.

  4. 4. Ask for shorter answers

    Output costs 4 times as much as input on DeepSeek V3, so a tight answer format (bullet points, JSON, a word limit) cuts the bill more than a shorter prompt.

Alternatives

Models priced closest to DeepSeek V3

From other providers, nearest on blended price. Quality and speed differ by task: test the shortlist on your own examples.

ModelProviderInput / 1MOutput / 1MBlended / 1MContext
DeepSeek V3DeepSeek$0.257$1.03$0.45164K
Mistral Small 3.1 24BMistral$0.351$0.555$0.402128K
Qwen3.8 27BQwen$0.035$1.50$0.4011M
Llama 4 MaverickMeta$0.35$1.00$0.513524K
SabaMistral$0.20$0.60$0.3033K
Llama 3.3 70B InstructMeta$0.71$0.71$0.71131K

DeepSeek

Other DeepSeek models

The same provider's newest models, for a quicker or more capable option.

ModelProviderInput / 1MOutput / 1MBlended / 1MContext
DeepSeek V4.1 FlashDeepSeek$0.05$0.199$0.0871M
DeepSeek V4 Pro 0423DeepSeek$0.90$1.50$1.051M
DeepSeek V4 Flash 0423DeepSeek$0.09$0.18$0.1131M

Questions about DeepSeek V3 pricing

How much does DeepSeek V3 cost per 1,000 tokens?

$0.0003 per 1,000 input tokens and $0.0010 per 1,000 output tokens, from the list price of $0.257 and $1.03 per million.

Is DeepSeek V3 cheaper than Mistral Small 3.1 24B?

No, on list price. DeepSeek V3 averages $0.45 per million tokens (3 parts input to 1 output) against $0.402 for Mistral Small 3.1 24B. A pricier model can still cost less overall if it needs fewer retries or shorter prompts.

What is DeepSeek V3's context window?

164K tokens: the most text the model reads and writes in one request. Long documents beyond that are split, or retrieved in parts with RAG.

How much would a chatbot cost on DeepSeek V3?

About $9.00 a month for 10,000 conversations a month, ~1,500 in and ~500 out each. Caching repeated instructions and routing simple questions to a cheaper model lower it further.

Does DeepSeek V3 support tool calling?

Yes. DeepSeek V3 can call functions and APIs you describe to it, which is what lets an AI agent look up data or take actions. It can also return answers that follow a JSON schema.

Can DeepSeek V3 read images or documents?

No. DeepSeek V3 reads text only; documents need their text extracted first.

How long can DeepSeek V3's answers be?

Up to 16K tokens per answer (about 12,000 words), within a 164K tokens context window.

How do I estimate the AI cost of my own product?

Count the requests a month and the tokens in and out of each, then multiply by the prices here. StackLift's free estimate does this for your exact scope, alongside the cost to build it.

Build with it

Scope a product on DeepSeek V3

Pick your features and integrations; the estimate returns the build cost, timeline and team, and the model's monthly running cost. Free, no signup.