Gemini 2.5 Pro Preview 06-05
| Released | 2025-06-05 |
|---|---|
| Price per 1M | $1.25 in · $10.00 out |
| Overall score | 49.7 · rank 116 |
| Context window | 1,048,576 tokens |
| Provider | |
| Weights | Proprietary |
Scores
| Dimension | Score | Rank | Benchmarks |
|---|---|---|---|
| Overall: Balanced | 49.7 | 116 | 22 |
| Overall: Coding | 49.5 | 114 | 12 |
| Overall: Reasoning | 49.3 | 119 | 10 |
| Overall: Agentic work | 48.4 | 125 | 12 |
| Overall: Frontend | 51.7 | 76 | 6 |
| Coding | 50.4 | 115 | 3 |
| Agents | 47.7 | 123 | 4 |
| Reasoning | 48.0 | 129 | 5 |
| Math | 49.3 | 119 | 3 |
| Knowledge | 51.9 | 82 | 2 |
| General | 50.1 | 107 | 2 |
| Multimodal | 55.7 | 42 | 3 |
| Value vs price | −1.7 | 141 |
Configurations
| Configuration | Balanced | Benchmarks |
|---|---|---|
| Unspecified setting | 50.2 ±1.3 | 21 |
| Thinking | 49.2 ±4.0 | 3 |
Evidence
Coding 50.4 ±3.5
| Benchmark | Result | Source | Date |
|---|---|---|---|
| GSO | 3.9% | Epoch AI | 2026-11-11 |
| Aider Polyglot | 83.1% | Epoch AI | 2026-11-11 |
| LiveCodeBench (Vals run) | 79.2% | Vals AI | 2026-09-01 |
Agents 47.7 ±3.0
| Benchmark | Result | Source | Date |
|---|---|---|---|
| METR task suite | 55.4% | Epoch AI | 2026-11-11 |
| Remote Labor Index | 0.8% | Epoch AI | 2026-11-11 |
| DeepResearch Bench | 42.8% | Epoch AI | 2026-11-11 |
| BALROG | 43.3% | Epoch AI | 2026-11-11 |
Reasoning 48.0 ±4.8
| Benchmark | Result | Source | Date |
|---|---|---|---|
| ARC-AGI-2 | 0.0% | Epoch AI | 2026-11-11 |
| ARC-AGI-1 | 33.0% | Epoch AI | 2026-11-11 |
| EnigmaEval | 5.6% | Epoch AI | 2026-11-11 |
| SimpleBench | 62.4% | Epoch AI | 2026-11-11 |
| GPQA Diamond | 84.9% | Epoch AI | 2025-06-05 |
| GPQA Diamond (Vals run) | 80.8% | Vals AI | 2026-09-01 |
Math 49.3 ±2.0
| Benchmark | Result | Source | Date |
|---|---|---|---|
| FrontierMath Tiers 1-3 (2025 set) | 10.3% | Epoch AI | 2025-07-03 |
| MATH Level 5 | 95.9% | Epoch AI | 2025-05-08 |
| FrontierMath Tier 4 (2025 set) | 2.1% | Epoch AI | 2025-07-03 |
Knowledge 51.9 ±1.3
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Humanity's Last Exam | 21.6% | Epoch AI | 2026-11-11 |
| MMLU-Pro (Vals run) | 84.1% | Vals AI | 2026-09-01 |
General 50.1 ±3.0
| Benchmark | Result | Source | Date |
|---|---|---|---|
| EQ-Bench Creative Writing v3 | 1,422 | EQ-Bench | 2026-11-11 |
| EQ-Bench Longform Creative Writing | 58.8% | EQ-Bench | 2026-11-11 |
Multimodal 55.7 ±9.9
| Benchmark | Result | Source | Date |
|---|---|---|---|
| VPCT | 48.0% | Epoch AI | 2026-11-11 |
| MMMU-Pro (Vals run) | 81.3% | Vals AI | 2026-09-01 |
| GeoBench | 86.0% | Epoch AI | 2026-11-11 |
Results not used in the score
| Benchmark | Result | Source | Why not used |
|---|---|---|---|
| ARC-AGI-1 | 31.3% | Epoch AI | A stronger configuration result was used |
| Aider Polyglot | 79.1% | Epoch AI | A stronger configuration result was used |
Names matched to this model
gemini-2-5-pro-exp-03-25, gemini-2-5-pro-preview, gemini-2-5-pro-preview-03-25, gemini-2-5-pro-preview-05-06, gemini-2-5-pro-preview-06-05