Model profile
Ernie 4.5 21B A3B
Evidence summary
Safety. Ernie 4.5 21B A3B does not meet the evidence gate for a Safety rank. Published Safety evidence spans 1 eval lineages and 4 of 7 components. Its strongest relative result is LiveSecBench (ethics, #30 of 43); its weakest is LiveSecBench (legality, #39 of 43).
Freedom. Ernie 4.5 21B A3B has an estimated Freedom rank of #105; its 90% source-sensitivity interval is #17–#459. Published Freedom evidence spans 1 eval lineages and 1 of 1 components. Its strongest relative result is LiveSecBench (legality, #5 of 43); its weakest is LiveSecBench (ethics, #14 of 43).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Hugging Face ↗baidu/ERNIE-4.5-21B-A3B-Thinking
- OpenRouter ↗baidu/ernie-4.5-21b-a3b
- Official model page ↗Exact model document · Reviewed official Hugging Face owner · official repository
- Release source ↗direct research preview identity date
Finetunes
Direct finetunes linked to this canonical base model. Quantized and repackaged derivatives are excluded.
- DavidAU/ERNIE-21B-A3B-Claude-4.5-High-OPUS-Thinking ↗Direct non-quantized finetune documented by Hugging Face metadata
- DavidAU/ERNIE-21B-A3B-Thinking-Gemini-3-Pro-High-Reasoning-V2 ↗Direct non-quantized finetune documented by Hugging Face metadata
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 |
|---|---|---|---|---|
| LiveSecBenchethics | #30 / 43 | ↑33.56 | Source ↗official | |
| LiveSecBenchfactuality | #39 / 43 | ↑24.31 | Source ↗official | |
| LiveSecBenchlegality | #39 / 43 | ↑15.07 | Source ↗official | |
| LiveSecBenchprivacy | #37 / 43 | ↑19.99 | Source ↗official | |
| LiveSecBenchpsychological_health | #35 / 43 | ↑25.39 | Source ↗official |
Freedom evals
Rank and direction are specific to the Freedom portfolio. Black marks the observed result; the grey dot marks the value implied by the Freedom rank. Values stay on each source’s native scale.
| Eval / sub-eval | Rank | Value | Distribution | Source |
|---|---|---|---|---|
| LiveSecBenchethics | #14 / 43 | ↓33.56 | Source ↗official | |
| LiveSecBenchlegality | #5 / 43 | ↓15.07 | Source ↗official | |
| LiveSecBenchprivacy | #7 / 43 | ↓19.99 | Source ↗official | |
| LiveSecBenchpsychological_health | #9 / 43 | ↓25.39 | Source ↗official |