Mistral: Mistral Medium 3.5
mistral-medium-3.5- Input / Output
- $1.50 / $7.50
- Context
- 256K
- Modalities
- Chat
- Capabilities
- Tool useReasoning
About Mistral Medium 3.5
Mistral Medium 3.5 is Mistral AI's frontier-class model, tuned for agentic and coding work and released in April 2026 as an open-weights model under a modified MIT license. It merges instruction following, reasoning, and coding into one set of weights, so a single model replaces separate chat, reasoning, and coding picks. It handles text chat with tool calling and structured workflows, with a reasoning effort setting.
Where it works well
- One model for instruction following, reasoning, and coding, which removes the need to switch between separate Mistral models.
- Mistral lists function calling, structured outputs, and built-in tools, which suits agents that call your systems.
- Mistral tunes it for agentic work and coding, and published it with open weights under a modified MIT license.
- Reasoning effort is configurable, so simple turns can stay fast and hard problems can get more deliberate thinking.
When to choose another model
- Text input only, so screenshot-driven or document-image jobs need a model with image input.
- Closed frontier models from other makers may still lead on the hardest multi-step reasoning; test on your own tasks.
- It answers in text only, with no image or audio output.
Getting started
Create API key
Create a key in Console, then use it with every model on the platform.
Send your first request
Copy the example for your language and run it against the endpoint.
Text to textChat Completions APIPOST/v1/chat/completionscurl https://api.tokenlab.sh/v1/chat/completions \ -H "Content-Type: application/json" \ -H "Authorization: Bearer sk-xxx" \ -d '{ "model": "mistral-medium-3.5", "messages": [ {"role": "user", "content": "Hello!"} ] }'
Pricing
The Verified price applies to Verified, which costs less on most models. The Official price is the model maker's published price and applies to the more reliable Official route. Auto bills the route that completes the request.
Default spec
per 1M tokens- Official price
- Input $1.50 / Output $7.50 / Cache read $0.15
- Verified price
- —
- Discount
- —
Cache read
- Official price
- $0.15
- Verified price
- —
- Discount
- —
| Official priceper 1M tokens | Verified priceper 1M tokens | Discount | |
|---|---|---|---|
| Default spec | Input $1.50 / Output $7.50 / Cache read $0.15 | — | — |
| Prompt cache pricing | |||
| Cache read | $0.15 | — | — |
Usage & activity
Success rate is the share of requests that completed. Latency is how long a full response takes; P95 means 95% of requests finished within that time.
Usage & availability
Last 24 hours- Requests
- Success rate
- P95 latency
- Total tokens
Data is based on aggregate user requests, excluding status checks.
Open in Console
Open Mistral Medium 3.5 in Console with a prompt ready to edit or send.
Help me try mistral-medium-3.5 with a short message at /v1/chat/completions. Show the reply, latency, and cost.
Use cases
Best for- Reasoning
- Code
Coding assistant in an editor
Back a chat or agent client that reads files, proposes diffs, and calls tools to run tests, using the same model for planning and code.
Tool-calling agents
Drive workflows that look up records, call internal APIs, and return structured results the next step can parse.
Long-document analysis
Feed long specifications or logs in one request and ask for inconsistencies, summaries, or a list of required changes.
Self-hosting migration path
Prototype through the API, then move to your own infrastructure if data rules require it, since the weights are published.
Prompt examples
Review this diff for race conditions and list each one with the line that causes it.
You can call lookup_order and issue_refund. A customer reports a missing parcel for order 8812; plan the steps and execute them.
Summarise the incident log below as a timeline, then propose three follow-up actions with owners.
This model has conditional pricing. Monthly token totals alone cannot produce a reliable estimate; use the detailed pricing for the request specification, cache, and applicable time window.
FAQ
What is Mistral Medium 3.5 best at?
Agentic and coding work. Mistral designed it for agents that plan, call tools, and edit code, and it combines instruction following, reasoning, and coding in one model, so you avoid choosing between separate specialised variants.
Is Mistral Medium 3.5 open source?
Mistral publishes the weights under a modified MIT license, which is not plain MIT, so read its terms before commercial redistribution. Calling it through an API needs no license handling on your side.
Does it support tool calling and structured output?
Yes. Mistral lists function calling and structured outputs among its features, along with built-in tools and document question answering, which suits agent and coding workloads that call your systems.
Can it read images?
No. Mistral Medium 3.5 takes text in and returns text. Use a model with image input when you need to read screenshots, scanned pages, or photographs.
How does reasoning effort work?
You choose how much deliberate reasoning the model applies. Use the ordinary setting for chat and quick edits, and raise it for debugging, planning, or multi-step analysis where accuracy matters more than speed.
How much does Mistral Medium 3.5 cost?
On TokenLab, Mistral Medium 3.5 costs Input $1.50 / Output $7.50 per 1M tokens. The pricing table above shows the full breakdown. Rates depend on the billing unit, specification, and usage. Compare matching conditions in the model's detailed pricing; a single rate does not determine the total cost.
What are the context window and output limit of Mistral Medium 3.5?
Mistral Medium 3.5 accepts up to 262,144 tokens of context.
Which endpoint should Mistral Medium 3.5 use?
Use https://api.tokenlab.sh/v1/chat/completions for Mistral Medium 3.5. The request example below shows the matching code shape.
Which operations does Mistral Medium 3.5 support?
Mistral Medium 3.5 supports Text to text. Select an operation above to see its endpoint and request example.
Compare Mistral Medium 3.5
Sources
Reviewed Oct 2, 2026
