Google
availableShows if the model has enough results for an index.Gemini 2.5 Pro
Gemini 2.5 Pro is a non-reasoning model from Google. 21 benchmarks count toward its score, in 6 categories.
IndexOverall score out of 100.46.2 ±5.0
CoverageShare of the index weight with results.85%
SpeedOutput tokens per second.87/s
Input / 1MUS dollars per 1M input tokens.$1.25 batch $0.625
Output / 1MUS dollars per 1M output tokens.$10 batch $5 US dollars per 1M output tokens in a batch.
ContextMaximum tokens in one request.1.05M
EloLMArena rating and rank.1458 (#36)
The index is a score out of 100. The ± range shows how much it can change. Batch work costs less.
122,554 votes. Elo shows what people prefer. It does not change the score.
CapabilitiesScore per category, out of 100.
Out of 100Results
21 counted| BenchmarkThe test name. | CategoryThe capability that the test measures. | ResultThe score from the publisher. | IndexThis result as a score out of 100. | RunThe settings of the run. | DateDate of the result. | Published byThe source of the result. |
|---|---|---|---|---|---|---|
| Artificial Analysis GPQA Diamond | Knowledge | 84.4% | 55.0 | — | — | Artificial Analysis |
| Graduate-Level Google-Proof Q&A | Knowledge | 83.0% | 54.9 | — | — | David Rein et al. |
| Artificial Analysis MMMU-Pro | Multimodal | 74.9% | 58.6 | — | — | Artificial Analysis |
| Artificial Analysis Long Context Reasoning | Reasoning | 69.0% | 56.0 | — | — | Artificial Analysis |
| Software Engineering Benchmark Verified | Coding | 63.8% | 48.7 | — | — | Carlos E. Jimenez et al. |
| SWE-bench | Coding | 54.4% | 41.2 | — | 1 Sept 2026 | Vals AI |
| τ²-Bench Tool-Agent-User Evaluation | Agentic | 54.1% | 37.6 | — | — | Victor Barres et al. |
| SWE-bench Verified | Coding | 53.6% | 40.5 | mini-SWE-agent | 26 Feb 2026 | SWE-bench team |
| SWE-bench Verified | Coding | 53.6% | 40.5 | mini-SWE-agent | 1 Sept 2026 | SWE-bench team |
| Artificial Analysis IFBench | Instruction | 48.7% | 38.5 | — | — | Artificial Analysis |
| Artificial Analysis SciCode | Coding | 46.3% | 57.1 | — | — | Artificial Analysis |
| Gert Labs Composite Game Benchmark | Agentic | 42.0% | 49.8 | — | — | Gert Labs |
| Terminal-Bench 1.0 | Agentic | 41.3% | 46.5 | — | 12 Jan 2026 | Vals AI |
| Artificial Analysis Omniscience Accuracy | Knowledge | 39.1% | 62.2 | — | — | Artificial Analysis |
| Artificial Analysis Coding Index | Coding | 33.3% | 42.4 | — | — | Artificial Analysis |
| Terminal-Bench 2.0 | Agentic | 30.3% | 44.3 | — | 4 Jun 2026 | Vals AI |
| Artificial Analysis Humanity's Last Exam | Knowledge | 22.5% | 49.0 | — | — | Artificial Analysis |
| Humanity's Last Exam | Knowledge | 18.8% | 44.7 | — | — | Center for AI Safety et al. |
| IOI v1 | Coding | 17.1% | 49.6 | — | 9 Aug 2026 | Vals AI |
| Artificial Analysis Intelligence Index | Knowledge | 16.1% | 42.5 | — | — | Artificial Analysis |
| τ²-bench Banking | Agentic | 13.7% | 8.6 | high effort · Sierra | 4 Aug 2026 | Sierra Research |
| Artificial Analysis Agentic Index | Agentic | 3.5% | 40.0 | — | — | Artificial Analysis |
| Critical Physics Tasks | Reasoning | 2.6% | 43.8 | — | — | Artificial Analysis |
| Vibe Code Bench v1.1 | Coding | 0.4% | 42.3 | OpenHands | 21 Sept 2026 | Vals AI |
| GDPval-AA normalized | Agentic | 0.0% | 35.6 | — | — | Artificial Analysis |
21 benchmarks count, from 25 of 25 results. A grey row does not count. Too few models took that benchmark.