Prices · DeepSeek
DeepSeek API pricing
DeepSeek through the OpenRouter endpoints llmwise uses costs from $0.30 to $0.40 per million input tokens, as of October 2026 (our catalog doesn't track DeepSeek's own API prices). Here's every DeepSeek model's price, what that means for a typical message, and what the same models cost per message in llmwise.
DeepSeek API prices
| Model | Input / output per 1M tokens | Typical message | Context window | In llmwise |
|---|---|---|---|---|
| DeepSeek V4 ProDeepSeek | $0.40 / $4.00 | $0.0044 | 1.05M tokens | 250/mo on Pro |
| DeepSeek V4.1 FlashDeepSeek | $0.30 / $1.20 | $0.0020 | 1.05M tokens | 60/day on Pro |
What a typical message costs
A chat message is rarely just your question: the model also reads the conversation so far, so 4,000 tokens in and 700 out is a fair middle. At these prices that's $0.0020 on DeepSeek V4.1 Flash, the least expensive DeepSeek model. A long document or a long chat can multiply the input; reasoning and long answers add output.
DeepSeek in llmwise
In llmwise you don't pay per token: you see how many messages you have left on each model before you send.
DeepSeek V4 Pro: 250 messages a month on Pro, and in Free's one-time trial of 5 messages.
DeepSeek V4.1 Flash: 60 messages a day on Pro, and in Free's one-time trial of 5 messages.
| Plan | Price | Everyday models | A month on the other models |
|---|---|---|---|
| Free | $0 | A one-time trial of 5 messages | The same 5, also on Claude Sonnet 5.5, Claude Sonnet 5, Claude Haiku 4.5, GPT-6.1 Sol, GPT-6 Sol, Gemini 3.1 Pro (preview), Gemini 3.8 Flash, DeepSeek V4 Pro, Grok 4.7, Kimi K3, GLM 5.3, and Mistral Large 4; one of them can be on Claude Opus 5.5 |
| Pro | $20 a month + applicable tax, shown before you pay | 60 a day | 31 messages on Claude Fable 5.1 or GPT-6 Astra; 62 on Claude Opus 5.5; 125 on Claude Sonnet 5.5-class models; or 250 on Claude Haiku 4.5-class models |
| Max | $50 a month + applicable tax, shown before you pay | 120 a day | 100 messages on Claude Fable 5.1 or GPT-6 Astra; 200 on Claude Opus 5.5; 400 on Claude Sonnet 5.5-class models; or 800 on Claude Haiku 4.5-class models |
| Ultra | $100 a month + applicable tax, shown before you pay | 200 a day | 225 messages on Claude Fable 5.1 or GPT-6 Astra; 450 on Claude Opus 5.5; 900 on Claude Sonnet 5.5-class models; or 1,800 on Claude Haiku 4.5-class models |
| Studio | $200 a month + applicable tax, shown before you pay | 200 a day | 500 messages on Claude Fable 5.1 or GPT-6 Astra; 1,000 on Claude Opus 5.5; 2,000 on Claude Sonnet 5.5-class models; or 4,000 on Claude Haiku 4.5-class models |
Every limit is published. Paid plans also have a monthly fair-use limit on AI cost: Pro $7.50, Max $20, Ultra $42, Studio $85. Using every message on your plan at typical sizes stays under it; very large messages and heavy research use it faster. Every limit, explained.
API or chat?
Building an app, a script or a pipeline that calls a model? You need an API: the provider's own, or a router in front of several. llmwise isn't one.
Want to ask questions, write, analyse files or code with a model yourself? A chat is simpler: no keys, no token maths, and in llmwise a fixed price per message you see before you send.
All prices side by side: LLM cost calculator.
Questions
How much does the DeepSeek API cost?
As of October 2026 (the price of the OpenRouter endpoints llmwise uses, not DeepSeek's own API), per million tokens in and out: DeepSeek V4 Pro $0.40 / $4.00; DeepSeek V4.1 Flash $0.30 / $1.20.
What does DeepSeek cost in llmwise?
Per message, not per token: DeepSeek V4 Pro, 250 messages a month on Pro; DeepSeek V4.1 Flash, 60 messages a day on Pro. You see how many are left before you send, however long the reply.
Can I use DeepSeek for free?
In llmwise, to try: DeepSeek V4 Pro and DeepSeek V4.1 Flash are on the Free plan, a one-time trial of 5 messages.
Claude, GPT, Gemini, DeepSeek, Grok, Kimi, GLM, and Mistral, in one chat.
See what a message costs before you send it. Free is 5 messages to try; sign in with an email link, no password or card.