Model profile
Qwen3.6 Max Preview
Evidence summary
Safety. Qwen3.6 Max Preview has an estimated Safety rank of #154; its 90% source-sensitivity interval is #73–#291. Its behavior-only rank is #148; company governance moves the combined estimate to #154. Published Safety evidence spans 5 eval lineages and 3 of 7 components. Its strongest relative result is AA-Omniscience (hallucination_rate, #68 of 345); its weakest is Every Model Cheats — Cybench Cheat Propensity (severe_anti_cheat_cheat_propensity_pct, #19 of 22).
Freedom. Qwen3.6 Max Preview does not meet the evidence gate for a Freedom rank.
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗qwen3-6-max
- OpenRouter ↗qwen/qwen3.6-max-preview
- Official model page ↗Exact model document · Alibaba Cloud · first party
Safety evals
Rank and direction are specific to the Safety portfolio. Black marks the observed result; the grey dot marks the value implied by the Safety rank. Values stay on each source’s native scale.
| Eval / sub-eval | Rank | Value | Distribution | Source |
|---|---|---|---|---|
| AA-Omnisciencehallucination_rate | #68 / 345 | ↓0.4619 | Source ↗official | |
| Arena Factuality — Text Arena (factuality-only weighting)factuality_bt_rating | #71 / 111 | ↑1428.0 | Source ↗official | |
| DystopiaBenchbasaglia_score | #36 / 50 | ↓68.63 | Source ↗official | |
| DystopiaBenchbaudrillard_score | #38 / 50 | ↓66.57 | Source ↗official | |
| DystopiaBenchhuxley_score | #38 / 50 | ↓75.8 | Source ↗official | |
| DystopiaBenchlaguardia_score | #39 / 50 | ↓69.4 | Source ↗official | |
| DystopiaBenchorwell_score | #34 / 50 | ↓73.17 | Source ↗official | |
| DystopiaBenchpetrov_score | #27 / 50 | ↓73.17 | Source ↗official | |
| Every Model Cheats — Cybench Cheat Propensitybaseline_cheat_propensity_pct | #15 / 22 | ↓39.1 | Source ↗official | |
| Every Model Cheats — Cybench Cheat Propensitysevere_anti_cheat_cheat_propensity_pct | #19 / 22 | ↓17.4 | Source ↗official | |
| Every Model Cheats — Cybench Cheat Propensitystandard_anti_cheat_cheat_propensity_pct | #17 / 22 | ↓26.1 | Source ↗official | |
| ODCV-Benchaverage_severity | #4 / 12 | ↓1.175 | Source ↗official | |
| ODCV-Benchmisalignment_rate | #5 / 12 | ↓28.75 | Source ↗official |
Freedom evals
No published sub-eval result contributes to this model’s Freedom profile.