Gemini 2.5 Pro Preview (2025-06-05)
Highlights
Top benchmark results for google/gemini-2-5-pro-preview-2025-06-05.
1063#3
0.82#3
0.88#12
0.41#11
0.05#8
12.38#6
4.04#24
1470#3
0.86#7
0.22#6
1470#1
1433#1
0.01#6
0.89#3
0.82#2
0.59#8
0.62#2
0.54#3
0.67#8
1.79#4
574#6
Benchmark table
Detailed scores across tracked benchmarks.
| Benchmark | Category | Top Score | Info | Self Reported | Source |
|---|---|---|---|---|---|
| Ai2 SciArena | - | 1063 | - | No | Source |
| Aider-Polyglot | code | 0.82 | Diff-Fenced | No | Source |
| AIME 2025 | math | 0.88 | Single Attempt | Yes | Source |
| ARC-AGI-1 | - | 0.41 | 16k Thinking | No | Source |
| ARC-AGI-2 | - | 0.05 | 32k Thinking | No | Source |
| Confabulations | - | 12.38 | - | No | Source |
| Elimation Game | - | 4.04 | - | No | Source |
| EQ-Bench 3 | - | 1470 | - | No | Source |
| GPQA Diamond | general-knowledge | 0.86 | Single Attempt | Yes | Source |
| Humanity's Last Exam | - | 0.22 | No Tools | Yes | Source |
| LMArena Text | - | 1470 | - | No | Source |
| LMArena WebDev | - | 1433 | 16th June 2025 | No | Source |
| MathArena Apex | - | 0.01 | - | No | Source |
| MMLU | - | 0.89 | - | Yes | Source |
| MMMU | - | 0.82 | Single Attempt | Yes | Source |
| NYT Connections | - | 0.59 | - | No | Source |
| SimpleBench | - | 0.62 | - | No | Source |
| SimpleQA | - | 0.54 | - | Yes | Source |
| SWE-Bench | code | 0.67 | Multiple Attempts | Yes | Source |
| Thematic Generalisation | - | 1.79 | - | No | Source |
| Vending Bench 2 | - | 574 | - | No | Source |
Benchmark comparisons
Use the selector to switch benchmarks and see how this model stacks up against its closest competitors.
NYT Connections
Compare this model with the leading peers for the selected benchmark.
Benchmark
0.59
Rank #8/41
41 models
Showing 11 models around the selected model (out of 41 total).