Model overview
Within the Mistral catalogue, mistral-medium-latest is the only model that carries reasoning, coding, agentic and multimodal capabilities at once, and the price reflects it: 1.5 USD per million input tokens against 7.5 for output, a five-to-one asymmetry no other model in this range applies. Compared with mistral-large-latest at 0.5 input and 1.5 output, you pay three times more to send tokens and five times more to receive them, so the tier only pays off on tasks where the extra reasoning changes the answer. Cached input falls to 0.15, exactly what the small tier charges at full rate, which makes long stable prompts far less punishing than long generations.
Billing conditions: Standard API price. Mistral publishes a 90% discount for cached input; cache-write prices are not published separately.
- Context window
- 0 tokens
- Knowledge cutoff
- —
- Price last verified
- First observed
Descriptions, capabilities, and tariffs come from the linked source. First-party prices take priority; OpenRouter fallbacks are labelled.