Gemini 3 Pro Preview
ModelActiveby Google · family “gemini-pro-preview”
Epoch Capabilities Index
Not scored by Epoch AI
Released
18 Nov 2025
Context window
Unknown
Max output
Unknown
Input $ / 1M tokens
Unknown
Output $ / 1M tokens
Unknown
Cached input $ / 1M
Unknown
Each benchmark shown separately with its own source
Benchmark results (21)
| Benchmark | Domain | Score | vs best recorded | Setting | Run | Source |
|---|---|---|---|---|---|---|
| CL-bench | long-context | 15.8% | 57% | — | — | External ↗ |
| ProofBench | math | 20.0% | 20% | — | — | External ↗ |
| Chess Puzzles | games | 31.0% ±4.6 | 43% | — | 8 Dec 2025 | Eval log ↗ |
| Remote Labor Index | agents | 1.3% | 6% | — | — | External ↗ |
| GDPval | agents | 40.3% | 81% | — | — | External ↗ |
| FrontierMath-Tier-4-2025-07-01-Privatesuperseded | math | 18.8% ±5.7 | 39% | — | 21 Nov 2025 | Epoch ↗ |
| GSO-Bench | coding | 18.6% | 40% | — | — | External ↗ |
| Terminal Bench | agents | 69.4% ±2.1 | 82% | agent: Ante | 18 Nov 2025 | External ↗ |
| ARC-AGI-2 | reasoning | 31.1% | 33% | — | — | External ↗ |
| METR Time Horizons | agents | 71.0% | 83% | — | — | External ↗ |
| GeoBench | multimodal | 84.0% | 95% | — | — | External ↗ |
| FrontierMath-2025-02-28-Privatesuperseded | math | 37.6% ±2.8 | 72% | — | 21 Nov 2025 | Epoch ↗ |
| VPCT | multimodal | 91.0% | 100% | — | — | External ↗ |
| HLE | knowledge | 37.5% | 68% | — | — | External ↗ |
| WeirdML | coding | 69.9% ±0.0 | 75% | — | — | External ↗ |
| OTIS Mock AIME 2024-2025 | math | 91.4% ±3.7 | 91% | — | 19 Nov 2025 | Eval log ↗ |
| Balrog | games | 58.1% | 85% | — | — | External ↗ |
| SimpleBench | reasoning | 76.4% | 93% | — | — | External ↗ |
| SWE-Bench verified | coding | 72.9% ±2.0 | 87% | — | 13 Feb 2026 | Eval log ↗ |
| GPQA diamond | science | 92.6% ±1.7 | 97% | — | 19 Nov 2025 | Eval log ↗ |
| ARC-AGI | reasoning | 75.0% | 76% | — | — | External ↗ |
Source: Epoch AI Benchmarking Hub (CC BY 4.0). “External” rows are leaderboard results Epoch collects from third parties. Best reported setting per benchmark is shown.
API list price over time
Price history
$ per 1M tokens
This model has no public API price listing
0 recorded prices on — (OpenRouter listing); re-read on every data refresh, most recently 1h ago. Steps show when the price changed.
Events
- Model launchMajorGoogle DeepMind releases Gemini 3 Pro PreviewEpoch AI Benchmarking Hub