Calibration
Claw-Eval → AA Agentic Index
AA Agentic Index is estimated from Claw-Eval with a inverse Michaelis–Menten curve fitted on 11 models measured on both: y = 0.43628·(x − 0.0000) / (0.0000 + 2.0000 − x), R² = 0.58, cross-validated error 5.1 pp. It is used for 18 estimates.
| Estimated model | Claw-Eval | AA Agentic Index | Source |
|---|---|---|---|
| Claude Opus 4.5 | 59.6% | 18.5% | estimated ± 5.1 pp, medium confidence |
| Claude Opus 4.6 | 70.4% | 23.7% | estimated ± 5.1 pp, medium confidence |
| Claude Sonnet 4.6 | 67.8% | 22.4% | estimated ± 5.1 pp, medium confidence |
| DeepSeek V3.2 | 40.2% | 11.0% | estimated ± 5.1 pp, medium confidence |
| Gemini 3 Flash | 49.2% | 14.2% | estimated ± 5.1 pp, medium confidence |
| GLM-5 | 57.7% | 17.7% | estimated ± 5.1 pp, medium confidence |
| GLM-5-Turbo | 55.8% | 16.9% | estimated ± 5.1 pp, medium confidence |
| GLM-5V-Turbo | 53.8% | 16.1% | estimated ± 5.1 pp, medium confidence |
| K-EXAONE 2.0 | 77.7% | 27.7% | estimated ± 5.1 pp, low confidence |
| LLaDA2.2-flash | 64.2% | 20.6% | estimated ± 5.1 pp, medium confidence |
| LLaDA2.2-mini | 57.2% | 17.5% | estimated ± 5.1 pp, medium confidence |
| MiMo-V2.5 | 62.3% | 19.7% | estimated ± 5.1 pp, medium confidence |
| MiMo-V2-Omni | 45.2% | 12.7% | estimated ± 5.1 pp, medium confidence |
| MiMo-V2-Pro | 57.8% | 17.7% | estimated ± 5.1 pp, medium confidence |
| Ornith-1.0-35B | 69.8% | 23.4% | estimated ± 5.1 pp, medium confidence |
| Ornith-1.0-397B | 77.1% | 27.4% | estimated ± 5.1 pp, low confidence |
| Ornith-1.0-9B | 63.1% | 20.1% | estimated ± 5.1 pp, medium confidence |
| Qwen3.5 397B | 56.8% | 17.3% | estimated ± 5.1 pp, medium confidence |