Mistral Medium 3.5 API Pricing

Mistral · released Apr 29, 2026 · 262K context · API model ID mistral-medium-2604

Mistral Medium 3.5 by Mistral API pricing: $1.50 per 1M input tokens and $7.50 per 1M output tokens (provider list price via models.dev, Oct 10, 2026).

Prices

Mistral Medium 3.5 prices, USD per 1M tokens
PriceUSD per 1M tokens
Input$1.50
Output$7.50
Cache read$0.15
Cache writeNot published

Mistral's list price for its own API, as compiled by models.dev and checked Oct 10, 2026, 09:00 UTC. Prices on cloud platforms and resellers can differ. Provider documentation

Estimated monthly cost

$1,575

per month

per day
$52.50
per request
$0.00525

Cost = input tokens × input price + output tokens × output price, 30 days a month. Cached input uses the cache read price. Cache write fees are not included.

Compare this workload across models

Monthly cost by workload

WorkloadPer month
Chatbot1,000 in / 500 out tokens, 10,000 requests a day$1,575/mo
RAG8,000 in / 500 out tokens, 5,000 requests a day$2,363/mo
Coding agent50,000 in / 2,000 out tokens, 1,000 requests a day, 80% cached input$1,080/mo

Context and capabilities

Context
262,144 tokens
Max output
262,144 tokens
Input
text, image
Output
text
Knowledge cutoff
Not published
Open weights
Yes
Tool calling
Yes
Reasoning
Yes
Structured outputs
Yes

Capabilities and limits as listed by models.dev.

Similarly priced models

Compare Mistral Medium 3.5

Price history

No price changes since we started tracking on Oct 8, 2026.

Questions

How much does Mistral Medium 3.5 cost per month?
For a chatbot handling 10,000 requests a day with 1,000 input and 500 output tokens each, Mistral Medium 3.5 costs about $1,575 per month at Mistral's list prices ($1.50 per 1M input tokens, $7.50 per 1M output tokens).
Does Mistral Medium 3.5 support prompt caching?
Yes. Cached input costs $0.15 per 1M tokens, compared with $1.50 for regular input.
How long is Mistral Medium 3.5's context window?
Mistral Medium 3.5 accepts up to 262,144 tokens (262K) of context and returns up to 262,144 tokens per response.