OverviewCompare models
Compare models
Two releases, one evaluator field and one scale. Missing results stay missing.
Compare one test
Choose a test and one evaluator field.
- 59.60%
- 19.70%
Bars start at zero. 2 reported results for the selected models; ranks apply to this selection only; ties share a rank. Published effort settings may vary. Source ↗
All shared tests
Same evaluator and unit. Percentage differences are percentage points.
Show all 57 shared tests Same evaluator · same unit
| Test / evaluator | Gemini 3.8 Flash | Claude Opus 5.5 | DifferenceModel 1 − model 2 |
|---|---|---|---|
| τ³-Bankingaa% · Higher is better | 44.95 | No result | — |
| ITBench-AAaa% · Higher is better | 52.54 | 38.23 | +14.31 pp |
| MLCR-AAaa% · Higher is better | 21.67 | 66.67 | -45.00 pp |
| ARC-AGI-3arc prize · standard% · Higher is better | 10.37 | No result | — |
| ARC-AGI-3arc prize · provider adapter% · Higher is better | 35.00 | No result | — |
| SWE-Bench Pro V2 Fullscale ai / seal% · Higher is better | 94.86 | No result | — |
| SWE-Bench Pro V2 HARDscale ai / seal% · Higher is better | 58.80 | No result | — |
| Vals RSI Indexvals% · Higher is better | 19.93 | 37.31 | -17.38 pp |
| Vibe Code Bench 1–100vals% · Higher is better | 18.77 | 30.36 | -11.59 pp |
| CUA-benchvals% · Higher is better | 4.17 | 14.00 | -9.83 pp |
| Harvey LAB-AAaa% · Higher is better | No result | 91.20 | — |
| AutomationBench-AAaa% · Higher is better | 59.93 | 69.54 | -9.61 pp |
| SciCodeaa% · Higher is better | 56.60 | 66.90 | -10.30 pp |
| GDP.pdfaa% · Higher is better | 21.00 | 26.20 | -5.20 pp |
| CritPtaa% · Higher is better | 18.29 | 31.71 | -13.42 pp |
| AA-Omniscienceaascore · Higher is better | 29.55 | 46.42 | -16.87 score |
| AA-LCR v1.1aa% · Higher is better | 81.33 | 84.67 | -3.34 pp |
| Terminal-Bench 4.0aa% · Higher is better | 19.70 | 59.60 | -39.90 pp |
| Terminal-Bench 4.0vals% · Higher is better | 19.19 | 65.15 | -45.96 pp |
| Terminal-Bench 2.1vals% · Higher is better | 81.27 | 87.64 | -6.37 pp |
| Terminal-Bench 2.1aa% · Higher is better | 87.64 | No result | — |
| SWE-bench (Verified)vals% · Higher is better | 80.00 | No result | — |
| LiveCodeBenchvals% · Higher is better | 89.48 | No result | — |
| GPQA Diamondvals% · Higher is better | 94.44 | No result | — |
| GPQA Diamondaa% · Higher is better | 95.25 | No result | — |
| MMLU Provals% · Higher is better | 90.22 | No result | — |
| MMMU Provals% · Higher is better | 89.08 | No result | — |
| MMMU Proaa% · Higher is better | 85.61 | 87.69 | -2.08 pp |
| Vibe Code Bench v1.1vals% · Higher is better | 78.65 | 90.29 | -11.64 pp |
| Terminal-Bench Science 0.1aa% · Higher is better | 10.00 | 59.05 | -49.05 pp |
| Terminal-Bench Science 0.1vals% · Higher is better | 8.57 | 47.14 | -38.57 pp |
| ProgramBenchvals% · Higher is better | 1.00 | 18.50 | -17.50 pp |
| ProofBench v1.1vals% · Higher is better | 48.00 | 100.00 | -52.00 pp |
| LegalBenchvals% · Higher is better | 86.99 | No result | — |
| Harvey's Legal Agent Benchmarkvals% · Higher is better | 10.00 | 3.75 | +6.25 pp |
| SkillsBenchvals% · Higher is better | 57.99 | No result | — |
| MysteryMechanismvals% · Higher is better | 36.49 | 49.55 | -13.06 pp |
| Vals Indexvals% · Higher is better | 54.83 | 66.97 | -12.14 pp |
| BioMysteryBenchvals% · Higher is better | 62.22 | 79.26 | -17.04 pp |
| Code Migrationvals% · Higher is better | 36.55 | 66.65 | -30.10 pp |
| CyberBench v1.1vals% · Higher is better | 43.75 | 55.36 | -11.61 pp |
| EMBvals% · Higher is better | 72.20 | 75.94 | -3.74 pp |
| Finance Agent (v2)vals% · Higher is better | 61.44 | 58.59 | +2.85 pp |
| IOIvals% · Higher is better | 56.94 | 95.06 | -38.12 pp |
| Legal Research Benchvals% · Higher is better | 38.94 | 50.48 | -11.54 pp |
| MedCodevals% · Higher is better | 48.13 | 49.80 | -1.67 pp |
| MedScribevals% · Higher is better | 84.50 | 91.43 | -6.93 pp |
| Public Benefits Bench v1.1vals% · Higher is better | 65.29 | 70.64 | -5.35 pp |
| SAGEvals% · Higher is better | 35.06 | 45.83 | -10.77 pp |
| SRE Benchvals% · Higher is better | No result | 33.59 | — |
| Tax Agent Benchvals% · Higher is better | 66.77 | 70.50 | -3.73 pp |
| DeepSWE v1.1datacurve% · Higher is better | 73.83 | No result | — |
| FrontierCode v1.1aa% · Higher is better | 36.36 | 52.17 | -15.81 pp |
| GDPval-AA v2.1aaElo · Higher is better | 1,411.95 | 1,846.17 | -434.22 Elo |
| AA-Briefcase v1.1aaElo · Higher is better | 1,201.59 | 1,821.85 | -620.26 Elo |
| Humanity's Last Examaa% · Higher is better | 47.82 | 61.35 | -13.53 pp |
| Artificial Analysis Intelligence Indexaaindex · Higher is better | 41 | 58 | -17.00 index |