Model
Inference & Serving
Mistral
1Efficient 2open-weight 3models 4from 5Mistral 6AI.
the six Ws · specification
W1
Who
Released by Mistral AI.
W2
What
Compact, efficient open models including Mixtral mixture-of-experts variants.
W3
Where
Weights run locally or via providers.
W4
When
When you want efficient open models.
W5
Why
Strong quality per parameter and token.
W6
With
A model runtime such as vLLM or Transformers.
W7
Watch
Apache-2.0 for open releases; weights on Hugging Face, some models are commercial.
alternatives
works with
MistralOpen-weightsEfficient