Question 1

What is Mistral Small 4 good for?

Accepted Answer

Mistral's smaller hybrid model unifying instruct, reasoning and coding (119 billion parameters, 6.5 billion active). 2 of 5 outputs load without errors; 1 more ran after one fix by the same model. Its best technical result was on the “Monitor test” prompt (98/100). Did not work (original output): Solar System, GTA-style game, Chess with an engine.

Question 2

How did Mistral Small 4 do in the AI Lab tests?

Accepted Answer

2 of 5 outputs load without errors; 1 more ran after one fix by the same model. Its best technical result was on the “Monitor test” prompt (98/100). Did not work (original output): Solar System, GTA-style game, Chess with an engine. In the objective perft chess test it passed 0 of 6 positions. Average time per prompt 2 min 01 s – rank 7 of 27 models by speed. Average cost per prompt $0.014 – rank 5 of 27 models from the cheapest. Overall score 21/100 – rank 25 of 27 models. Technical health (speed, errors, accessibility, mobile) 71/100 – rank 27 of 27 models. Automatic UX rating 52/100 – rank 22 of 27 models.

Question 3

How much does Mistral Small 4 cost?

Accepted Answer

Through the API Mistral Small 4 costs $0.15 per million input and $0.6 per million output tokens (price list as of 17/10/2026).

Question 4

When was Mistral Small 4 released?

Accepted Answer

Mistral Small 4 by Mistral was released on 16/03/2026.

Mistral Small 4

What is Mistral Small 4 good for?

How did Mistral Small 4 do in the AI Lab tests?

Outputs

Solar System

GTA-style game

Monitor test

New York & Central Park map

Chess with an engine