Magistral Small
Everyday reasoning, document analysis, and long code reviews on a budget.
What are Magistral Small's specs and price?
Magistral Small, built by Mistral, ships a 128K-token context window and a 33K-token max output, released 2026-09. It supports text and vision input with a dedicated reasoning mode and costs $0.75 per million blended tokens, the 16th-cheapest of 49 models we track.
What are Magistral Small's specs?
| Context window | 128K tokens |
| Max output | 33K tokens |
| Modalities | text, vision |
| Extended thinking | Yes |
| Released | 2026-09 |
| Knowledge cutoff | Not published |
| Provider | Mistral |
Verified 2026-10-01 — source.
Where does Magistral Small rank?
What are Magistral Small's strengths?
- Open-weights reasoning with image analysis
- Runs on a laptop-class machine
- Budget reasoning at $0.50/$1.50 per million tokens
What else should you know about Magistral Small?
What are common questions about Magistral Small?
What is Magistral Small's context window?
Magistral Small has a 128K-token context window and a 33K-token max output — the 48th-largest context of the 49 current models we track. Source: https://docs.mistral.ai/models/magistral-small-1-2-25-09, verified 2026-10-01.
Does Magistral Small support vision or audio input?
Yes — Magistral Small accepts vision input in addition to text.
Does Magistral Small have a reasoning or extended-thinking mode?
Yes — Magistral Small exposes a dedicated reasoning mode for multi-step problems.
When was Magistral Small released, and what is its knowledge cutoff?
Magistral Small was released 2026-09.
How much does Magistral Small cost, and who provides it?
Magistral Small is served by Mistral at $0.75/M blended tokens (3:1 input:output) — the 16th-cheapest of 49 current models. Full pricing breakdown: /llm-api-pricing/magistral-small.