Gemini 2.5 Flash (Sep 2025)
| Released | 2025-09-25 |
|---|---|
| Price per 1M | Not listed |
| Overall score | 48.6 · rank 126 |
| Provider | |
| Weights | Proprietary |
Scores
| Dimension | Score | Rank | Benchmarks |
|---|---|---|---|
| Overall: Balanced | 48.6 | 126 | 8 |
| Overall: Coding | 49.0 | 117 | 3 |
| Overall: Reasoning | 50.9 | 105 | 2 |
| Overall: Agentic work | 45.9 | 145 | 3 |
| Overall: Frontend | 51.1 | 80 | 4 |
| Coding | 50.9 | 111 | 1 |
| Agents | 42.4 | 149 | 1 |
| Reasoning | 51.9 | 100 | 1 |
| Knowledge | 48.8 | 98 | 1 |
| General | 48.9 | 119 | 1 |
| Multimodal | 51.8 | 66 | 3 |
Configurations
| Configuration | Balanced | Benchmarks |
|---|---|---|
| Unspecified setting | 48.4 ±2.2 | 8 |
| Thinking | 50.1 ±4.0 | 4 |
Evidence
Coding 50.9 ±5.3
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LiveCodeBench (Vals run) | 76.2% | Vals AI | 2026-09-01 |
Agents 42.4 ±2.5
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Terminal-Bench 2.0 | 17.1% | Epoch AI | 2025-11-03 |
Reasoning 51.9 ±4.8
| Benchmark | Result | Source | Date |
|---|---|---|---|
| GPQA Diamond (Vals run) | 81.6% | Vals AI | 2026-09-01 |
Knowledge 48.8 ±4.7
| Benchmark | Result | Source | Date |
|---|---|---|---|
| MMLU-Pro (Vals run) | 83.7% | Vals AI | 2026-09-01 |
General 48.9 ±4.1
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LMArena Text | 1,403 | LMArena | 2026-11-09 |
Multimodal 51.8 ±2.9
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LMArena Vision | 1,229 | LMArena | 2026-11-09 |
| MMMU-Pro (Vals run) | 80.8% | Vals AI | 2026-09-01 |
| VPCT | 46.2% | Epoch AI | 2026-11-11 |
Results not used in the score
| Benchmark | Result | Source | Why not used |
|---|---|---|---|
| GPQA Diamond (Vals run) | 76.5% | Vals AI | A stronger configuration result was used |
| LiveCodeBench (Vals run) | 75.1% | Vals AI | A stronger configuration result was used |
| MMLU-Pro (Vals run) | 83.7% | Vals AI | A stronger configuration result was used |
| MMMU-Pro (Vals run) | 79.5% | Vals AI | A stronger configuration result was used |
| Terminal-Bench 2.0 | 16.9% | Epoch AI | A stronger configuration result was used |
| Terminal-Bench 2.0 | 16.4% | Epoch AI | A stronger configuration result was used |
| Terminal-Bench 2.0 | 15.4% | Epoch AI | A stronger configuration result was used |
Names matched to this model
gemini-2-5-flash-preview-09-2025, gemini-2-5-flash-sep-2025