Model profile
Mistral Medium 3.5
Evidence summary
Mistral Medium 3.5 has an estimated overall rank of #252; its 90% source-sensitivity interval is #120–#297. Its behavior-only rank is #243; company governance moves the combined estimate to #252. Published evidence spans 4 evals and 4 of 7 behavior components. Its strongest relative result is PHARE (bias_resistance_diagnostic, #11 of 66); its weakest is DystopiaBench (petrov_score, #50 of 50).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗mistral-medium-3-5
- Hugging Face ↗mistralai/Mistral-Medium-3.5-128B
- NVIDIA ↗qc69jvmznzxy/mistral-medium-3.5-128b
- OpenRouter ↗mistralai/mistral-medium-3-5
- Official model page ↗Exact model document · Reviewed official Hugging Face owner · official repository
- Release source ↗repository created at proxy
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 |
|---|---|---|---|---|
| AA-Omnisciencehallucination_rate | #163 / 328 | ↓0.8159 | Source ↗official | |
| Arena Factuality — Text Arena (factuality-only weighting)factuality_bt_rating | #62 / 112 | ↑1433.0 | Source ↗official | |
| DystopiaBenchbasaglia_score | #50 / 50 | ↓77.7 | Source ↗official | |
| DystopiaBenchbaudrillard_score | #48 / 50 | ↓79.27 | Source ↗official | |
| DystopiaBenchhuxley_score | #50 / 50 | ↓86.8 | Source ↗official | |
| DystopiaBenchlaguardia_score | #50 / 50 | ↓76.57 | Source ↗official | |
| DystopiaBenchorwell_score | #50 / 50 | ↓82.27 | Source ↗official | |
| DystopiaBenchpetrov_score | #50 / 50 | ↓90.97 | Source ↗official | |
| PHAREbias_resistance_diagnostic | #11 / 66 | ↑0.6098 | Source ↗official | |
| PHAREhallucination_resistance_diagnostic | #54 / 70 | ↑0.677 | Source ↗official | |
| PHAREharm_resistance_diagnostic | #46 / 70 | ↑0.9115 | Source ↗official | |
| PHAREjailbreak_resistance_diagnostic | #55 / 67 | ↑0.3746 | Source ↗official |
Values evaluations
Descriptive values and political-framing results are separate from safety/ethics ranks. Each strip shows the evaluation’s observed model range; its endpoint labels state what lower and higher values mean.