Prices · Per message
What an AI message costs, model by model
Pro ($20 a month) gives up to 60 everyday messages a day and, each month, up to 31 messages on Claude Fable 5.1 or GPT-6 Astra; 62 on Claude Opus 5.5 or Gemini 4 Argon; 125 on Claude Sonnet 5.5-class models; or 250 on Claude Haiku 4.5-class models. In llmwise you see how many messages you have left on each model before you send; here's that count for every model, how far each plan goes, and what the same message costs at API prices.
What a message costs on each model
Every model's count in llmwise, the most messages first, next to what a typical message would cost through its API.
| Model | On Pro | On Free | Typical message at API prices |
|---|---|---|---|
| Claude Haiku 5.5Anthropic | 60 a day | Yes | $0.0008 |
| GPT-6 LunaOpenAI | 60 a day | No | $0.0008 |
| GLM 5.3 FlashZ.ai | 60 a day | No | $0.0010 |
| DeepSeek V4.1 FlashDeepSeek | 60 a day | No | $0.0020 |
| Mistral Large 4Mistral | 250 a month | No | $0.0042 |
| DeepSeek V4 ProDeepSeek | 250 a month | No | $0.0044 |
| Gemini 3.8 FlashGoogle | 250 a month | No | $0.0056 |
| Claude Haiku 4.5Anthropic | 250 a month | No | $0.0075 |
| GLM 5.3Z.ai | 250 a month | No | $0.0087 |
| Grok 4.7xAI | 250 a month | No | $0.0122 |
| Claude Sonnet 5.5Anthropic | 125 a month | No | $0.0150 |
| Claude Sonnet 5Anthropic | 125 a month | No | $0.0150 |
| GPT-6.1 SolOpenAI | 125 a month | No | $0.0150 |
| GPT-6 SolOpenAI | 125 a month | No | $0.0150 |
| Gemini 3.1 Pro (preview)Google | 125 a month | No | $0.0164 |
| Kimi K3Moonshot | 125 a month | No | $0.0225 |
| Claude Opus 5.5Anthropic | 62 a month | No | $0.0300 |
| Gemini 4 ArgonGoogle | 62 a month | No | $0.0300 |
| Claude Fable 5.1Anthropic | 31 a month | No | $0.0750 |
| GPT-6 AstraOpenAI | 31 a month | No | $0.0750 |
How far each plan goes
Each paid plan has a daily count for the everyday models and one monthly allowance for the rest (and the paid tools), shown as messages per model. Free is a one-time trial.
| Plan | Price | Everyday models | On Claude Fable 5.1 or GPT-6 Astra | On Claude Opus 5.5 or Gemini 4 Argon | On Claude Sonnet 5.5-class models | On Claude Haiku 4.5-class models |
|---|---|---|---|---|---|---|
| Free | $0 | In the 5 | Not on this plan | Not on this plan | Not on this plan | Not on this plan |
| Plus | $10 | 30 a day | Not on this plan | 31 | 62 | 125 |
| Pro | $20 | 60 a day | 31 | 62 | 125 | 250 |
| Max | $50 | 120 a day | 100 | 200 | 400 | 800 |
| Ultra | $100 | 200 a day | 225 | 450 | 900 | 1,800 |
| Studio | $200 | 200 a day | 500 | 1,000 | 2,000 | 4,000 |
Every limit is published. Paid plans also have a monthly fair-use limit on AI cost: Plus $3.75, 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.
What changes the price, and what doesn't
The reply's length, its reasoning and its tool steps don't change a message's price: it's fixed when you send.
Long chats count more: once the conversation sent to the model passes 64k tokens, each reply counts as 2; past 128k, as 4; past 200k the chat is full. The composer says so before you send: “Long chat — each reply counts as 2. Start a new chat to save messages.”
On Free, web search works inside your 5 trial messages; on a paid plan each search counts as one more message of the model in use.
Paid tools add their own fixed price: a code run counts as 1 Claude Haiku 4.5 message, an image from 2 Claude Haiku 4.5 messages an image (by image model), a backend app 5 Claude Haiku 4.5 messages per started 10 minutes, a research report 15 Claude Haiku 4.5 messages (Standard) or 75 Claude Haiku 4.5 messages (Deep). Documents, slides, sheets and downloads are free.
A reply cut off because a model's provider failed costs nothing.
Per message or per token?
Through an API you pay per token. At the prices in our catalog as of October 2026, a typical message costs from $0.0008 on Claude Haiku 5.5 to $0.0750 on GPT-6 Astra, and a long document or a long chat multiplies that. In llmwise you pay a monthly plan instead, and see how many messages it has left on each model.
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.
Work out API costs for your own usage with the LLM cost calculator, or see every plan on the pricing page.
Questions
What does one message cost in llmwise?
An everyday message on Claude Haiku 5.5, GPT-6 Luna, DeepSeek V4.1 Flash, and GLM 5.3 Flash, or a share of the monthly allowance on the others, shown as messages left before you send. That's exactly what's counted, however long the reply.
How many messages does Pro include?
Pro ($20 a month + applicable tax, shown before you pay) has 60 everyday messages a day and, each month, 31 messages on Claude Fable 5.1 or GPT-6 Astra; 62 on Claude Opus 5.5 or Gemini 4 Argon; 125 on Claude Sonnet 5.5-class models; or 250 on Claude Haiku 4.5-class models, or any mix.
Is a plan cheaper than paying per token?
It depends on how much you use. Through an API you pay for exactly the tokens each message uses; a plan is a fixed monthly price with a fixed price per message, so you know what you'll pay before you send. If you'd rather not do token maths, a plan is simpler; if you build software, you need an API anyway.
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.