Model profile / Google
Google's preview model for advanced multimodal understanding and agentic workloads.
Four areas of evidence
Mathematical and scientific problem solving.
Score uses the admitted direct native evidence under the selected method.
Software engineering and programming.
Score uses the admitted direct native evidence under the selected method.
Factual reliability and instruction following.
Score uses the admitted direct native evidence under the selected method.
Tool use, workflows and professional tasks.
Score uses the admitted direct native evidence under the selected method.
Current index snapshot · 2026-09-16. Each category uses its own scale; scores across categories are not directly comparable. Evidence counts describe the retained Capabilities inputs; a family can contain several tasks or configurations. A point lead does not establish superiority.
Performance in context
Gemini 3.1 Pro Preview is highlighted wherever a compatible measurement is available.
Lumina Capabilities Index
The current Capabilities cohort, with this model highlighted.
Chart loads as you explore
Lumina Capabilities Index
The current Capabilities cohort, with this model highlighted.
Chart loads as you explore
Explore the detail
Score uses the admitted direct native evidence under the selected method.
Score uses the admitted direct native evidence under the selected method.
Score uses the admitted direct native evidence under the selected method.
Score uses the admitted direct native evidence under the selected method.
Score uses the admitted direct native evidence under the selected method.
| Benchmark | Result | Published configuration | Source |
|---|---|---|---|
| FrontierMath v2 Tiers 1-3v2 / task implementation 2.0.0 | 59.649% | Gemini 3.1 Pro PreviewEpoch AI benchmark runner | Epoch AI permanent refresh source ↗Observed 2026-06-11 · checked 2026-08-17source-checked · excluded |
| FrontierMath v2 Tier 4v2 / task implementation 2.0.0 | 26.829% | Gemini 3.1 Pro PreviewEpoch AI benchmark runner | Epoch AI permanent refresh source ↗Observed 2026-06-11 · checked 2026-08-17source-checked · excluded |
| GPQA Diamond1.0.11 | 94.444% | Gemini 3.1 Pro Preview (high)Epoch AI benchmark runner | Epoch AI permanent refresh source ↗Observed 2026-08-06 · checked 2026-08-17source-checked · excluded |
| GPQA Diamond1.0.5 | 94.1% | Gemini 3.1 Pro PreviewEpoch AI benchmark runner | Epoch AI permanent refresh source ↗Observed 2026-02-20 · checked 2026-08-17source-checked · excluded |
| SWE-Bench ProSource-native version | 54.2% | Exact model variant and provider evaluation configuration stated in the Gemini 3.5 Flash model card; single-attempt where specified.Provider evaluation harness | Gemini 3.5 Flash model card ↗Observed 2026-05-19 · checked 2026-07-15provider-reported · ranking-eligible |
| MCP AtlasSource-native version | 78.2% | Exact model variant and provider evaluation configuration stated in the Gemini 3.5 Flash model card; single-attempt where specified.Provider evaluation harness | Gemini 3.5 Flash model card ↗Observed 2026-05-19 · checked 2026-07-15provider-reported · ranking-eligible |
| OSWorld-VerifiedSource-native version | 76.2% | Exact model variant and provider evaluation configuration stated in the Gemini 3.5 Flash model card; single-attempt where specified.Provider evaluation harness | Gemini 3.5 Flash model card ↗Observed 2026-05-19 · checked 2026-07-15provider-reported · ranking-eligible |
| Finance Agent v2Source-native version | 43% | Exact model variant and provider evaluation configuration stated in the Gemini 3.5 Flash model card; single-attempt where specified.Provider evaluation harness | Gemini 3.5 Flash model card ↗Observed 2026-05-19 · checked 2026-07-15provider-reported · reference-only |
| CharXivSource-native version | 83.3% | Exact model variant and provider evaluation configuration stated in the Gemini 3.5 Flash model card; single-attempt where specified.Provider evaluation harness | Gemini 3.5 Flash model card ↗Observed 2026-05-19 · checked 2026-07-15provider-reported · ranking-eligible |
| MMMU-ProSource-native version | 80.5% | Exact model variant and provider evaluation configuration stated in the Gemini 3.5 Flash model card; single-attempt where specified.Provider evaluation harness | Gemini 3.5 Flash model card ↗Observed 2026-05-19 · checked 2026-07-15provider-reported · ranking-eligible |
