Overview

Mistral’s large models are optimized for efficient, high-quality text generation.

Inputs and Outputs

Text/JSON structured outputs.

Inference & Decoding

Standard decoding and structured outputs.

Safety & Compliance

Apply model and external guardrails.

Evals & Quality

Competitive on public benchmarks; verify on internal datasets.

Deployment & Ops

Available across multiple clouds and inference runtimes.

Pricing & Licensing

Commercial terms; compare per-token pricing.

References

See YAML references.