Model profile
Mistral Medium
Mistral AIdeveloper
2023-12-11release date
#149 / 312overall rank
4eval lineages
Evidence summary
Mistral Medium has an estimated overall rank of #149; its 90% source-sensitivity interval is #65–#302. Its behavior-only rank is #141; company governance moves the combined estimate to #149. Published evidence spans 4 evals and 5 of 7 behavior components. Its strongest relative result is OR-Bench (over_refusal_rate, #6 of 25); its weakest is OR-Bench (toxic_acceptance_rate, #24 of 25).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗mistral-medium
- OpenRouter ↗mistralai/mistral-medium
- Official model documentation ↗Family-level model document · mistral · first party
- Release source ↗curated outward evidence
Published eval results
Rank is within that sub-eval. Black marks the observed result; the grey dot marks the value implied by the global rank. Values stay on each source’s native scale.
| Eval / sub-eval | Rank | Value | Distribution | Source |
|---|---|---|---|---|
| BlueBench AttaQ-100attaq_harmlessness_reward_pct | #10 / 18 | ↑84.68 | Source ↗official | |
| MASKlying_probability_pct | #50 / 53 | ↓57.4 | Source ↗official | |
| OR-Benchover_refusal_rate | #6 / 25 | ↓14 | Source ↗official | |
| OR-Benchtoxic_acceptance_rate | #24 / 25 | ↓22.5 | Source ↗official | |
| PacifAIstp_score | #4 / 7 | ↑84.62 | Source ↗official |