Overall score
62.6
Rank #17 · A confidence
Google · stable
Google's stable multimodal model for agentic, coding and high-throughput workloads.
Overall score
62.6
Rank #17 · A confidence
Capability
60.0
33 eligible benchmark families
Coverage
100%
Confidence A
Benchmarks scored
39
54 raw result rows
Model specification
Identity, configuration and price are preserved as dated source records.
Category profile
A category remains Not evaluated until an eligible source-qualified result is attached.
Benchmark scores
Best published score per benchmark. Open any row for the full per-benchmark leaderboard.
| Benchmark | Category | Score | Date |
|---|---|---|---|
| τ²-Bench TelecomRanking weighted | agents | 95.3216% | 2026-07-15 |
| τ-benchRanking weighted | agents | 95.3% | 2026-07-15 |
| GPQA DiamondRanking weighted | reasoning | 92.2% | 2026-07-15 |
| CharXivRanking weighted | multimodal | 84.2% | 2026-05-19 |
| MMMU-ProRanking weighted | multimodal | 83.87% | 2026-07-15 |
| MCP AtlasRanking weighted | agents | 83.6% | 2026-05-19 |
| OSWorld-VerifiedRanking weighted | agents | 78.4% | 2026-05-19 |
| LongBench v2 | long context | 77.3% | 2026-07-15 |
| IFBench | instruction following | 76.3% | 2026-07-15 |
| IFEval | instruction following | 76.3% | 2026-07-15 |
| Terminal-BenchRanking weighted | agents | 76.2% | 2026-05-19 |
| BenchLM Multimodal priorRanking weighted | multimodal | 75.6% | 2026-07-15 |
| LiveBenchRanking weighted | reasoning | 75.02% | 2026-07-15 |
| BenchLM Reasoning priorRanking weighted | reasoning | 72.7% | 2026-07-15 |
| ARC-AGI-2Ranking weighted | reasoning | 72.1% | 2026-05-19 |
| BenchLM Knowledge priorRanking weighted | research | 71.9% | 2026-07-15 |
| AA Long Context ReasoningRanking weighted | research | 69.3333% | 2026-07-15 |
| BenchLM Coding priorRanking weighted | coding | 65.4% | 2026-07-15 |
| BenchLM Math priorRanking weighted | mathematics | 61.7% | 2026-07-15 |
| Finance Agent v2 | research | 57.9% | 2026-05-19 |
| ToolathlonRanking weighted | agents | 56.5% | 2026-05-19 |
| SWE-Bench ProRanking weighted | coding | 55.1% | 2026-05-19 |
| SciCodeRanking weighted | coding | 53.1% | 2026-07-15 |
| EnterpriseOps-Gym-AARanking weighted | agents | 50.0746% | 2026-07-15 |
| BenchLM Agentic priorRanking weighted | agents | 49.1% | 2026-07-15 |
| CursorBenchRanking weighted | coding | 48.8% | 2026-07-15 |
| APEX-Agents-AARanking weighted | agents | 47.0501% | 2026-07-15 |
| AA AutomationBenchRanking weighted | agents | 42.6% | 2026-07-15 |
| GDPval-AA v2Ranking weighted | agents | 42.442% | 2026-07-15 |
| Terminal-Bench HardRanking weighted | agents | 40.9091% | 2026-07-15 |
| Long-Horizon-Terminal-BenchRanking weighted | agents | 40.9% | 2026-07-15 |
| ITBench-AARanking weighted | agents | 40.3484% | 2026-07-15 |
| Humanity's Last ExamRanking weighted | reasoning | 40.2% | 2026-05-19 |
| FrontierMathRanking weighted | mathematics | 38.966% | 2026-07-15 |
| Blueprint-Bench 2 | multimodal | 33.6% | 2026-05-19 |
| τ³-BankingRanking weighted | agents | 25.3608% | 2026-07-15 |
| AA-OmniscienceRanking weighted | research | 22.6833 index | 2026-07-15 |
| CritPtRanking weighted | reasoning | 13.1429% | 2026-07-15 |
| Harvey LAB-AARanking weighted | agents | 1.6667% | 2026-07-15 |
Showing the best of 54 raw result rows across 39 benchmark families.
Price and efficiency
Operational values are shown only when the source publishes a comparable measurement or a valid derivation.
Verification summary
News and X posts are structurally excluded from this model's evidence graph.
Missing publication dimensions: value. The last source check for this exact model was 2026-07-14.
Related models
Exact variants remain separate even when they share a provider or family.
Gemini 3.1 · preview
Gemini 3.1 Flash-LiteGemini 3.1 · stable
Gemini 3 ProGemini 3 · stable