Model profile
Qwen3.5 397B A17B
Evidence summary
Qwen3.5 397B A17B has an estimated overall rank of #42; its 90% source-sensitivity interval is #33–#133. Its behavior-only rank is #35; company governance moves the combined estimate to #42. Published evidence spans 8 evals and 7 of 7 behavior components. Its strongest relative result is LiveSecBench (ethics, #2 of 43); its weakest is Enkrypt AI Safety Leaderboard (cbrn_attack_non_success_rate, #233 of 260).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗qwen3-5-397b-a17b
- Hugging Face ↗Qwen/Qwen3.5-397B-A17B
- OpenRouter ↗qwen/qwen3.5-397b-a17b
- Official model page ↗Exact model document · Reviewed official Hugging Face owner · official repository
- Release source ↗direct research preview identity date
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 | Better | Source |
|---|---|---|---|---|---|
| AA-Omnisciencehallucination_rate | #147 / 311 | 0.7985 | ↓ lower | Source ↗official | |
| BullshitBench v2clear_pushback_rate | #9 / 105 | 0.735 | ↑ higher | Source ↗official | |
| Enkrypt AI Safety Leaderboardbias_attack_non_success_rate | #19 / 260 | 50.39 | ↑ higher | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #233 / 260 | 75.5 | ↑ higher | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #29 / 260 | 97.22 | ↑ higher | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #99 / 258 | 97.45 | ↑ higher | Source ↗official | |
| Inkling-Small model card — FORTRESSbenign_answer_rate | #6 / 10 | 95.4 | ↑ higher | Source ↗official | |
| Inkling-Small model card — FORTRESSharmful_refusal_rate | #6 / 10 | 77.3 | ↑ higher | Source ↗official | |
| Inkling-Small model card — StrongREJECTsafety_rate | #1 / 10 | 99.4 | ↑ higher | Source ↗official | |
| LiveSecBenchethics | #2 / 43 | 89.63 | ↑ higher | Source ↗official | |
| LiveSecBenchfactuality | #8 / 43 | 72.1 | ↑ higher | Source ↗official | |
| LiveSecBenchlegality | #16 / 43 | 62.2 | ↑ higher | Source ↗official | |
| LiveSecBenchprivacy | #2 / 43 | 94.2 | ↑ higher | Source ↗official | |
| LiveSecBenchpsychological_health | #3 / 43 | 89.45 | ↑ higher | Source ↗official | |
| SABERoverall_safety_rate | #5 / 13 | 26.58 | ↑ higher | Source ↗official | |
| SABERscenario_a_safety_rate | #5 / 13 | 30.62 | ↑ higher | Source ↗official | |
| SABERscenario_b_safety_rate | #4 / 13 | 36 | ↑ higher | Source ↗official | |
| SABERscenario_c_safety_rate | #7 / 13 | 15.02 | ↑ higher | Source ↗official | |
| SM-Benchadversarial | #42 / 73 | 80.49 | ↑ higher | Source ↗official | |
| SM-Benchambiguous_interpretation | #54 / 73 | 79.76 | ↑ higher | Source ↗official | |
| SM-Benchanti_hallucination | #12 / 73 | 98.43 | ↑ higher | Source ↗official | |
| SM-Bencheq_boundaries | #50 / 73 | 58.43 | ↑ higher | Source ↗official | |
| SM-Benchoverfit | #52 / 73 | 56.83 | ↑ higher | Source ↗official | |
| TACbase_welfare_rate | #31 / 68 | 28.21 | ↑ higher | Source ↗self run |
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.
UGI Political Values
Agent-ValueBench Moral Foundations (MFT08)
Agent-ValueBench HEXACO
| Dimension | Value | Distribution |
|---|---|---|
| Openness to experience | 6.5 | |
| Honesty-humility | 5.1 | |
| Extraversion | 6.3 | |
| Agreeableness | 6.7 | |
| Conscientiousness | 7.8 |