Model profile
Step 3.5 Flash
Evidence summary
Step 3.5 Flash has an estimated overall rank of #203; its 90% source-sensitivity interval is #93–#249. Its behavior-only rank is #204; company governance moves the combined estimate to #203. Published evidence spans 4 evals and 5 of 7 behavior components. Its strongest relative result is LiveSecBench (factuality, #12 of 43); its weakest is TAC (base_welfare_rate, #60 of 68).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗step-3-5-flash
- Hugging Face ↗stepfun-ai/Step-3.5-Flash
- NVIDIA ↗qc69jvmznzxy/step-3.5-flash
- OpenRouter ↗stepfun/step-3.5-flash
- 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 | Better | Source |
|---|---|---|---|---|---|
| AA-Omnisciencehallucination_rate | #235 / 311 | 0.884 | ↓ lower | Source ↗official | |
| BullshitBench v2clear_pushback_rate | #70 / 105 | 0.175 | ↑ higher | Source ↗official | |
| LiveSecBenchethics | #16 / 43 | 56.5 | ↑ higher | Source ↗official | |
| LiveSecBenchfactuality | #12 / 43 | 62.37 | ↑ higher | Source ↗official | |
| LiveSecBenchlegality | #31 / 43 | 26.76 | ↑ higher | Source ↗official | |
| LiveSecBenchprivacy | #23 / 43 | 46.86 | ↑ higher | Source ↗official | |
| LiveSecBenchpsychological_health | #12 / 43 | 68.03 | ↑ higher | Source ↗official | |
| TACbase_welfare_rate | #60 / 68 | 18.59 | ↑ 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.