Overall score
54.8
Rank #42 · A confidence
Google · preview
Google's preview model for advanced multimodal understanding and agentic workloads.
Overall score
54.8
Rank #42 · A confidence
Capability
58.0
30 eligible benchmark families
Coverage
92%
Confidence A
Benchmarks scored
31
39 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 | 95.614% | 2026-07-15 |
| τ-benchRanking weighted | agents | 95.6% | 2026-07-15 |
| GPQA DiamondRanking weighted | reasoning | 94.14% | 2026-07-15 |
| MMMU-ProRanking weighted | multimodal | 83.9% | 2026-07-15 |
| CharXivRanking weighted | multimodal | 83.3% | 2026-05-19 |
| LiveBenchRanking weighted | reasoning | 79.93% | 2026-07-15 |
| MCP AtlasRanking weighted | agents | 78.2% | 2026-05-19 |
| ARC-AGI-2Ranking weighted | reasoning | 77.1% | 2026-05-19 |
| IFBench | instruction following | 77.1% | 2026-07-15 |
| IFEval | instruction following | 77.1% | 2026-07-15 |
| OSWorld-VerifiedRanking weighted | agents | 76.2% | 2026-05-19 |
| Terminal-BenchRanking weighted | agents | 73.78% | 2026-07-15 |
| AA Long Context ReasoningRanking weighted | research | 72.6667% | 2026-07-15 |
| SciCodeRanking weighted | coding | 58.9% | 2026-07-15 |
| SWE-Bench ProRanking weighted | coding | 54.2% | 2026-05-19 |
| Long-Horizon-Terminal-BenchRanking weighted | agents | 53.8% | 2026-07-15 |
| Terminal-Bench HardRanking weighted | agents | 53.7879% | 2026-07-15 |
| GAIARanking weighted | agents | 46.1% | 2026-07-15 |
| Humanity's Last ExamRanking weighted | reasoning | 44.72% | 2026-07-15 |
| Finance Agent v2 | research | 43% | 2026-05-19 |
| EnterpriseOps-Gym-AARanking weighted | agents | 42.1665% | 2026-07-15 |
| AA AutomationBenchRanking weighted | agents | 37.5% | 2026-07-15 |
| FrontierMathRanking weighted | mathematics | 36.9% | 2026-07-15 |
| AA-OmniscienceRanking weighted | research | 32.9333 index | 2026-07-15 |
| APEX-Agents-AARanking weighted | agents | 32.0059% | 2026-07-15 |
| ITBench-AARanking weighted | agents | 30.3309% | 2026-07-15 |
| Blueprint-Bench 2 | multimodal | 26.5% | 2026-05-19 |
| GDPval-AA v2Ranking weighted | agents | 23.1155% | 2026-07-15 |
| CritPtRanking weighted | reasoning | 17.7143% | 2026-07-15 |
| τ³-BankingRanking weighted | agents | 16.4948% | 2026-07-15 |
| Harvey LAB-AARanking weighted | agents | 0% | 2026-07-15 |
Showing the best of 39 raw result rows across 31 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.
Capability ranking is complete for the current snapshot. 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 3.5 · stable
Gemini 3.1 Flash-LiteGemini 3.1 · stable
Gemini 3 ProGemini 3 · stable