Gemini 3.0 Pro Preview
Highlights
Top benchmark results for google/gemini-3-0-pro-preview-2025-11-18.
1#1
0.88#1
0.45#1
0.81#1
0.93#1
0.94#1
0.46#1
2439#1
0.23#1
0.92#1
0.81#1
0.12#1
0.26#1
0.07#1
0.75#1
0.72#1
0.76#2
0.85#1
0.54#2
5478#1
0.88#1
Benchmark table
Detailed scores across tracked benchmarks.
| Benchmark | Category | Top Score | Info | Self Reported | Source |
|---|---|---|---|---|---|
| AIME 2025 | math | 1 | With Code Execution | Yes | Source |
| ARC-AGI-1 | - | 0.88 | Deep Think | No | Source |
| ARC-AGI-2 | - | 0.45 | Deep Think, With Tools | Yes | Source |
| CharXiv-Reasoning | - | 0.81 | - | Yes | Source |
| Global PICA | - | 0.93 | - | Yes | Source |
| GPQA Diamond | general-knowledge | 0.94 | Deep Think, Tools Off | Yes | Source |
| Humanity's Last Exam | - | 0.46 | With Search & Code Execution | Yes | Source |
| LiveCodeBench Pro | - | 2439 | - | Yes | Source |
| MathArena Apex | - | 0.23 | - | No | Source |
| MMMLU | - | 0.92 | - | Yes | Source |
| MMMU Pro | - | 0.81 | - | Yes | Source |
| OmniDocBench 1.5 | - | 0.12 | - | Yes | Source |
| OpenAI MRCR 8 Needle 128k | - | 0.77 | - | Yes | Source |
| OpenAI MRCR 8 Needle 1m | - | 0.26 | - | Yes | Source |
| ScreenSpot-Pro | - | 0.07 | - | Yes | Source |
| SimpleBench | - | 0.75 | - | No | Source |
| SimpleQA | - | 0.72 | - | Yes | Source |
| SWE-Bench | code | 0.76 | Single Attempt | Yes | Source |
| Tau2 Bench | - | 0.85 | - | Yes | Source |
| Terminal Bench 2.0 | - | 0.54 | - | Yes | Source |
| Vending Bench 2 | - | 5478 | - | Yes | Source |
| Video MMMU | - | 0.88 | - | Yes | Source |
Benchmark comparisons
Use the selector to switch benchmarks and see how this model stacks up against its closest competitors.
OpenAI MRCR 8 Needle 1m
Compare this model with the leading peers for the selected benchmark.
Benchmark
0.26
Rank #1/1
1 models
Showing 1 models around the selected model (out of 1 total).