Gemini 2.5 Flash (Sep 2025)
Gemini 2.5 Flash (Sep 2025) (Google DeepMind) was released on Sep 25, 2025. In the Epoch AI capability ranking (ECI), it ranks #106 out of 267 models, with 143.00. On Arena, which measures human preference in blind comparisons, it has 1,450 ELO points (#126). Against models of the same generation, its best result is on The Agent Company: 51.8%, #2 of 6.
Where Gemini 2.5 Flash (Sep 2025) stands out and where it falls behind, by the numbers
Gemini 2.5 Flash (Sep 2025) has published results on 24 tests compiled by Epoch AI. The yardstick here is its generation: its position among the models released between Dec 29, 2024 and Dec 24, 2025 that took the same test. Against the whole archive, which includes 2023 models, a recent release lands near the top of almost everything, and that says nothing.
Where it stands out
Tests where it sits in the top third of its own generation.
Where it falls behind
Tests where it sits in the bottom half of its own generation.
See all 24 tests, one by one
| Test | Score | In its generation | Overall | Test record |
|---|---|---|---|---|
| AIME 2024 and 2025 (OTIS) | 73.1% | #22 / 56 | #93 / 168 | 100.0% GPT-5.5 |
| Dtbench | 76.5% | #22 / 47 | #82 / 148 | 98.9% Claude Opus 5.5 |
| Lmca | 27.49 | #19 / 41 | #81 / 121 | 68.23 Claude Opus 5.5 |
| CritPt | 1.1% | #11 / 32 (tied with 1) | #77 / 112 | 32.3% GPT-5.6 Sol |
| WeirdML | 41.9% | #24 / 41 | #68 / 112 | 93.6% GPT-6 Astra |
| ARC-AGI 2 | 2.5% | #23 / 36 | #66 / 83 | 95.0% GPT-6 Astra |
| ALE-Bench | 662 | #15 / 25 | #65 / 82 | 2,951 GPT-6 Astra |
| ARC-AGI | 33.3% | #22 / 35 | #65 / 81 | 98.5% Claude Fable 5 |
| SimpleBench | 41.2% | #18 / 31 | #49 / 74 | 81.9% Claude Fable 5 |
| FrontierMath | 4.8% | #22 / 35 | #45 / 65 | 51.7% GPT-5.5 |
| ForecastBench | 60.60 | #15 / 28 | #22 / 62 | 62.50 o3 |
| Vending Bench 2 | 549 | #9 / 14 | #48 / 57 | 15,515 GPT-6 Astra |
| FrontierMath tier 4 | 4.2% | #8 / 26 (tied with 7) | #23 / 48 | 39.6% GPT-5.5 Pro |
| Hle | 12.1% | #12 / 21 | #24 / 39 | 54.8% GPT-6 Astra |
| Aider Polyglot | 55.10 | #17 / 32 (tied with 1) | #18 / 38 | 88.00 GPT-5 |
| Creative writing (Lech Mazur) | 7.65 | #19 / 28 | #21 / 38 | 8.60 GPT-5 |
| EnigmaEval | 2.7% | #17 / 21 | #28 / 36 | 39.3% Claude Fable 5 |
| Terminal-Bench | 17.1% | #18 / 19 | #31 / 33 | 84.7% GPT-5.5 |
| BALROG | 33.5% | #7 / 11 | #13 / 32 | 68.3% GPT-6 Astra |
| Fiction.liveBench | 68.8% | #7 / 29 (tied with 1) | #8 / 31 | 100.0% o3 |
| GeoBench | 3,854 | #8 / 16 | #10 / 24 | 4,333 Gemini 3 Flash |
| VPCT | 46.2% | #9 / 18 | #9 / 22 | 91.0% Gemini 3 Pro |
| The Agent Company | 51.8% | #2 / 6 | #2 / 14 | 52.4% DeepSeek-V3.2-Exp |
| Spatialviz Bench | 36.9% | #2 / 5 | #3 / 7 | 44.7% Gemini 2.5 Pro (Jun 2025) |
For each test, the model's best effort level counts, as in the ranking. Source of the scores: Epoch AI, under CC BY 4.0.
What to compare Gemini 2.5 Flash (Sep 2025) with
The models from other labs closest in capability and released in the same period, at most two per lab. Click any of them to see its page.
| Model | Lab | ECI | ELO | LiveBench | Release |
|---|---|---|---|---|---|
| GLM-4.7 | Z.ai (Zhipu AI) | 143.79 | 1,495 | n/a | Dec 22, 2025 |
| Qwen 3.6 Flash | Alibaba | 143.05 | n/a | n/a | Apr 27, 2026 |
| Gemini 2.5 Flash (Sep 2025) | Google DeepMind | 143.00 | 1,450 | n/a | Sep 25, 2025 |
| Claude Opus 4 | Anthropic | 142.65 | 1,442 | n/a | May 22, 2025 |
| Qwen3.5-35B-A3B | Alibaba | 142.48 | 1,457 | n/a | Feb 24, 2026 |
| GPT-5.5 Instant | OpenAI | 142.45 | 1,456 | n/a | May 5, 2026 |
| Claude Haiku 4.5 | Anthropic | 142.40 | 1,454 | n/a | Oct 15, 2025 |
Gemini 2.5 Flash (Sep 2025) or Qwen 3.6 Flash?
Qwen 3.6 Flash is the model from another lab closest to Gemini 2.5 Flash (Sep 2025) in capability. Across the 7 tests both took, this is the score:
| Test | Gemini 2.5 Flash (Sep 2025) | Qwen 3.6 Flash |
|---|---|---|
| FrontierMath tier 4 | 4.2% | 0.0% |
| FrontierMath | 4.8% | 10.3% |
| ALE-Bench | 662 | 326 |
| SimpleBench | 41.2% | 35.2% |
| AIME 2024 and 2025 (OTIS) | 73.1% | 84.4% |
| Lmca | 27.49 | 30.98 |
| Dtbench | 76.5% | 77.1% |
Unweighted score: each test counts one point, whether math or coding. The dot marks who did better on each row.
Other models from Google DeepMind
The models from the same lab closest in time, newest first.
| Model | Release | ECI | ELO | LiveBench | Access |
|---|---|---|---|---|---|
| Gemma 4 31B IT | Apr 2, 2026 | 142.86 | n/a | n/a | Open weights (restricted use) |
| Gemini 3.1 Pro | Feb 19, 2026 | 154.67 | 1,528 | 77.0 | API access |
| Gemini 3 Flash | Dec 17, 2025 | 151.80 | 1,511 | n/a | API access |
| Gemini 3 Pro | Nov 18, 2025 | 152.90 | 1,523 | n/a | API access |
| Gemini 2.5 Flash (Sep 2025) | Sep 25, 2025 | 143.00 | 1,450 | n/a | API access |
| Gemini 2.5 Flash-Lite (Jun 2025) | Jun 17, 2025 | 133.95 | n/a | n/a | API access |
| gemini-2.5-flash-lite-preview-06-17 (Default thinking length) | Jun 17, 2025 | 133.92 | 1,401 | n/a | API access |
| Gemini 2.5 Pro (Jun 2025) | Jun 5, 2025 | 145.31 | 1,507 | n/a | API access |
| Gemini 2.5 Pro Exp (Mar 2025) | Mar 25, 2025 | 144.20 | n/a | n/a | API access |
Have you used Gemini 2.5 Flash (Sep 2025)? Rate it
Users' opinion, not measurement: the rating is kept apart from the ranking and does not move anyone's position. One vote per browser, no sign-up, and you can change your rating later.
What people ask about Gemini 2.5 Flash (Sep 2025)
Where does Gemini 2.5 Flash (Sep 2025) rank among AI models?
In the ROO3 AI Benchmark, ordered by the Epoch AI ECI, Gemini 2.5 Flash (Sep 2025) ranks #106 out of 267 models, with an ECI of 143.00. The ECI is a composite index that combines several hard benchmarks into a single scale. The position changes when a new model comes out, and the page is refreshed six times a day.
Is Gemini 2.5 Flash (Sep 2025) better than Qwen 3.6 Flash?
Across the 7 tests both models took and Epoch AI compiled, Gemini 2.5 Flash (Sep 2025) comes out ahead on 3, behind on 4 and tied on 0. It is a test-by-test count with no weighting: a win in math counts the same as a win in coding. To choose between the two, look at the tests that resemble your own use.
What is Gemini 2.5 Flash (Sep 2025) best at?
Against models of the same generation (released up to 9 months before and up to 3 months after it), the best relative results of Gemini 2.5 Flash (Sep 2025) are: The Agent Company (#2 of 6); Fiction.liveBench (#7 of 29); CritPt (#11 of 32). The yardstick is its position among the models of that period that took the same test.
Where do the numbers on this page come from?
From the same three sources as the ranking, all openly licensed: Epoch AI (ECI and individual tests, CC BY 4.0), Arena (human-preference ELO, CC BY 4.0) and LiveBench (overall average, CC BY-SA 4.0). ROO3 does not measure any model. When the page shows a figure published by the maker itself, it appears in the written analysis, labeled as the maker's, and never enters the ranking.
Where each number comes from
Epoch AI, "AI Benchmarking Hub". Published at epoch.ai. Data under the CC BY 4.0 licence, reordered by roo3.co.
The Arena (LMArena) leaderboard, from the public lmarena-ai/leaderboard-dataset on Hugging Face, under the CC BY 4.0 licence. We are not affiliated with the Arena.
LiveBench, by the LiveBench team, published at livebench.ai under the CC BY-SA 4.0 license. The overall average is recomputed by roo3.co with the same formula as the website (average of the categories), and those numbers stay under CC BY-SA 4.0. We are not affiliated with LiveBench.
← See the full ranking