Model overview
Largest of the three lightweight variants and still flat-rated at 0.2 per million tokens either way, ministral-14b-latest costs twice the 0.1 of the smallest step and a third more than the 0.15 of the middle one. The comparison worth running is with mistral-small-latest: input is more expensive here, 0.2 against 0.15, but output is 0.2 against 0.6, three times less, so the arithmetic flips as soon as answers grow longer than prompts. What you give up is the multimodal and multilingual support declared on that tier, since the lightweight line lists only agentic and lightweight capabilities. Cached input follows the same 90 percent discount, at 0.02.
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.