Skip to main content
Mistral AI

Mistral Medium 3.5 pricing, context window and limits

Mistral Medium 3.5 costs $1.5 per million input tokens and $7.5 per million output tokens at Mistral AI list price. It accepts up to 262,144 tokens of context and returns up to 262,144 tokens.

  • Model id mistral-medium-2604
  • Snapshot

What does Mistral Medium 3.5 cost?​

List prices in US dollars per million tokens
Input$1.5
Output$7.5
Cached input (read)$0.15
Cache write—

US dollars per million tokens. A dash means no price is published.

How much can Mistral Medium 3.5 read and write?​

Token limits
Context window262,144
Max output262,144

What can Mistral Medium 3.5 do?​

Capabilities
Vision (image input)Yes
Tool callingYes
Structured outputYes
ReasoningYes
Inputtext, image
Outputtext

What does a request to Mistral Medium 3.5 cost?​

Cost of example requests at list price, in US dollars
A typical request: 1,000 tokens in, 500 out$0.0053
A long request: 10,000 tokens in, 1,000 out$0.022
1,000 typical requests$5.25
1 million tokens in and 1 million out$9

Input read from the provider's prompt cache costs $0.15 per million tokens instead of $1.5, 90% less. A request that repeats a 1,000-token prompt from the cache costs $0.0039 instead of $0.0053.

Among 40 Mistral AI models with a published price, Mistral Medium 3.5 is the 35th cheapest by input price. Its context window is the 5th largest of 40. List prices only; discounts, batch prices and prices above 200K tokens are not shown.

How do I call Mistral Medium 3.5 through DVARA?​

Configure the mistral provider in the DVARA LLM Gateway and send the model id mistral-medium-2604. DVARA, the AI gateway, applies your policies, audit and cost attribution to the call. See LLM cost attribution and budgets.

In DVARA, if you haven't set your own price for a model, its calls are costed at the price in the catalogue that ships with your DVARA version. A price you set in Flightdeck always wins. You can switch catalogue prices off. Then a model with no price you set gets no cost record: its calls count against no budget and appear in no report, and the gateway records the model as unpriced. Every cost record says which price it used.

Which other Mistral AI models are there?​

All 42 Mistral AI models · Search the full catalogue

Which models from other providers are similar to Mistral Medium 3.5?​

These are the closest in context window and input price, one per provider.

Snapshot 16 October 2026. DVARA reviews every price weekly before it is published. Get the whole catalogue as llm-models.json. Prices are list prices in US dollars per million tokens; check your provider contract before you budget.