TokenBingeby JoDevelop
All model prices
Mistral AI

mistral-small-latest

Official price

Model overview

Multimodal and multilingual work starts here: 0.15 per million input tokens and 0.6 for output make mistral-small-latest the least expensive Mistral model that accepts images. Against mistral-large-latest, the gap is 3.3 times on input (0.5) and 2.5 times on output (1.5) for an identical declared pair of multimodal and multilingual capabilities, which is the argument for starting at this tier and moving up only when quality forces it. The comparison with ministral-8b-latest is sharper: the same 0.15 to send, but 0.6 against 0.15 to receive, four times more, so answer length rather than prompt size decides between the two.

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.

Capabilities and formats

  • text
  • agentic
  • multimodal
  • multilingual

Performance references

No comparable benchmark data is currently available.

PRICE EVOLUTION

Price history

Original observation and every later change, including any change of price source.

  1. Original observed priceOfficial price
    Input: $0.15Cached input: $0.015Cache write 5m: Cache write 1h: Output: $0.6