Other · model
Sage-A1.5 benchmark scores
As of 2026-10-29, Sage-A1.5 (Other) has measured scores on 10 benchmarks and estimated scores on 76 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 92.1% | estimated ± 10.9 pp, low confidence |
| AA-HLE | 45.4% | estimated ± 4.2 pp, high confidence |
| AA-SciCode | 44.1% | estimated ± 3.0 pp, high confidence |
| CritPt | 20.0% | measured |
| GDPval-AA | 43.0% | estimated ± 7.7 pp, medium confidence |
| GPQA Diamond | 92.5% | estimated ± 3.4 pp, high confidence |
| HLE | 50.9% | estimated ± 9.9 pp, medium confidence |
| IFBench | 81.6% | measured |
| AA-LCR | 81.0% | measured |
| LiveCodeBench v6 | 90.0% | estimated ± 2.7 pp, high confidence |
| SciCode | 44.3% | measured |
| BrowseComp | 81.7% | measured |
| HLE w/ tools | 55.1% | measured |
| IFEval | 94.9% | estimated ± 3.6 pp, high confidence |
| LongBench v2 | 62.3% | estimated ± 5.8 pp, low confidence |
| VITA-Bench | 45.4% | measured |
| τ²-bench results | 97.8% | estimated ± 3.1 pp, high confidence |
| AA Coding Index | 67.0% | estimated ± 7.5 pp, medium confidence |
| APEX-Agents-AA | 28.7% | estimated ± 4.2 pp, high confidence |
| BioMysteryBench (human-difficult) | 48.1% | estimated ± 6.8 pp, low confidence |
| DeepSearchQA | 86.4% | estimated ± 7.3 pp, low confidence |
| SWE-bench Verified | 81.9% | estimated ± 3.7 pp, high confidence |
| AutomationBench | 30.9% | estimated ± 10.5 pp, low confidence |
| BFCL v4 | 66.7% | estimated ± 2.9 pp, medium confidence |
| CyberGym | 74.3% | estimated ± 7.3 pp, medium confidence |
| JobBench | 41.1% | estimated ± 10.1 pp, low confidence |
| SWE-bench Pro | 61.2% | measured |
| WideResearch | 78.6% | estimated ± 3.5 pp, high confidence |
| AA Agentic Index | 36.0% | estimated ± 8.2 pp, medium confidence |
| AA-IFBench | 80.4% | estimated ± 7.4 pp, medium confidence |
| Artificial Analysis Intelligence Index | 39.1% | estimated ± 8.5 pp, medium confidence |
| GPQA | 92.5% | estimated ± 11.6 pp, low confidence |
| AA-Omniscience Accuracy | 46.0% | estimated ± 8.7 pp, medium confidence |
| MMMLU | 90.8% | estimated ± 0.7 pp, low confidence |
| AA MMLU-Pro | 88.4% | estimated ± 1.5 pp, low confidence |
| Gert Labs | 64.1% | estimated ± 7.4 pp, medium confidence |
| AA EnterpriseOps-Gym | 43.5% | estimated ± 3.3 pp, medium confidence |
| AA Harvey LAB v1.0 | 89.5% | estimated ± 1.7 pp, medium confidence |
| AA Tau3 Banking | 36.6% | estimated ± 3.9 pp, high confidence |
| ARC-AGI-1 | 89.7% | estimated ± 7.8 pp, low confidence |
| CursorBench 3.1 | 55.4% | estimated ± 5.6 pp, low confidence |
| CursorBench 3.2 | 59.6% | estimated ± 3.6 pp, high confidence |
| FrontierCode 1.1 Main | 40.7% | estimated ± 5.5 pp, medium confidence |
| FrontierSWE v2 | 17.3% | estimated ± 9.0 pp, medium confidence |
| MazeBench | 0.0% | estimated ± 6.1 pp, low confidence |
| OSWorld-Verified | 74.7% | estimated ± 5.8 pp, medium confidence |
| Vals GPQA Diamond | 91.5% | estimated ± 8.4 pp, medium confidence |
| Vals LiveCodeBench | 84.9% | estimated ± 0.8 pp, medium confidence |
| Vals MMLU-Pro | 88.4% | estimated ± 2.6 pp, medium confidence |
| Vals SWE-bench | 70.1% | estimated ± 5.3 pp, low confidence |
| AA AutomationBench | 50.0% | estimated ± 11.6 pp, low confidence |
| CWE-Bench-AA | 56.1% | estimated ± 5.2 pp, medium confidence |
| AA Cyber Index | 36.5% | estimated ± 12.8 pp, low confidence |
| DeepsecBench-AA | 19.4% | estimated ± 7.4 pp, medium confidence |
| GDP.pdf | 13.7% | estimated ± 4.9 pp, medium confidence |
| AA ITBench | 46.5% | estimated ± 4.1 pp, high confidence |
| CursorBench 4.0 | 36.4% | estimated ± 3.6 pp, medium confidence |
| FrontierCode 2.0 | 20.6% | estimated ± 7.1 pp, low confidence |
| GraphWalks BFS 256K–1M | 71.9% | estimated ± 14.8 pp, low confidence |
| HLE w/o tools | 44.6% | estimated ± 6.9 pp, medium confidence |
| OSWorld 2.1 | 60.5% | estimated ± 8.3 pp, low confidence |
| PostTrainBench v1.1 | 26.4% | estimated ± 4.7 pp, high confidence |
| SWE Multilingual | 72.7% | measured |
| Toolathlon-Verified | 70.3% | estimated ± 5.7 pp, medium confidence |
| Hermes Bench | 67.2% | estimated ± 4.8 pp, low confidence |
| Hermes Index | 34.4% | estimated ± 10.0 pp, low confidence |
| ExploitGym | 6.7% | estimated ± 9.2 pp, low confidence |
| Claw-Eval | 62.9% | estimated ± 7.1 pp, low confidence |
| DeepPlanning | 58.5% | estimated ± 9.1 pp, low confidence |
| INCLUDE | 84.4% | estimated ± 2.0 pp, medium confidence |
| MCP Atlas | 77.9% | estimated ± 10.9 pp, low confidence |
| MMLU-ProX | 84.1% | estimated ± 1.3 pp, high confidence |
| NL2Repo | 48.4% | estimated ± 6.1 pp, medium confidence |
| QwenClawBench | 61.8% | estimated ± 2.0 pp, medium confidence |
| Toolathlon | 51.3% | estimated ± 4.2 pp, high confidence |
| HealthBench Hard | 32.5% | estimated ± 8.6 pp, low confidence |
| LiveCodeBench Pro | 84.7% | estimated ± 6.0 pp, low confidence |
| ResearchClawBench | 18.2% | estimated ± 2.3 pp, medium confidence |
| BioMysteryBench (human-solvable) | 89.1% | estimated ± 0.6 pp, medium confidence |
| FrontierCode 1.1 Extended | 56.4% | estimated ± 3.0 pp, low confidence |
| STEMBench | 2.4% | estimated ± 0.2 pp, high confidence |
| Agents' Last Exam | 30.0% | estimated ± 13.0 pp, low confidence |
| OpenHarmony Bench | 55.0% | estimated ± 4.9 pp, low confidence |
| MRCRv2 | 85.8% | estimated ± 11.3 pp, low confidence |
| MMLongBench-Doc | 61.5% | estimated ± 2.6 pp, low confidence |
| PinchBench | 90.0% | measured |