Gemini 3.5 Flash
Gemini 3.5 Flash (Google DeepMind) was released on May 19, 2026. In the Epoch AI capability ranking (ECI), it ranks #37 out of 267 models, with 154.38. On Arena, which measures human preference in blind comparisons, it has 1,526 ELO points (#27). On LiveBench, the overall average is 74.6, #31 out of 58 scored models. Against models of the same generation, its best result is on SWE-bench Verified: 79.3%, #3 of 23.
Where Gemini 3.5 Flash stands out and where it falls behind, by the numbers
Gemini 3.5 Flash has published results on 31 tests compiled by Epoch AI. The yardstick here is its generation: its position among the models released between Aug 22, 2025 and Aug 17, 2026 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 31 tests, one by one
| Test | Score | In its generation | Overall | Test record |
|---|---|---|---|---|
| GPQA Diamond | 92.8% | #14 / 71 | #20 / 173 | 96.4% Gemini 4 Argon |
| AIME 2024 and 2025 (OTIS) | 95.6% | #21 / 71 (tied with 3) | #37 / 168 | 100.0% GPT-5.5 |
| Dtbench | 94.7% | #12 / 67 (tied with 1) | #19 / 148 | 98.9% Claude Opus 5.5 |
| Lmca | 47.15 | #20 / 65 | #27 / 121 | 68.23 Claude Opus 5.5 |
| CritPt | 13.1% | #25 / 64 | #42 / 112 | 32.3% GPT-5.6 Sol |
| WeirdML | 62.6% | #23 / 57 | #26 / 112 | 93.6% GPT-6 Astra |
| SciCode | 53.9% | #17 / 60 | #33 / 104 | 66.9% Claude Opus 5.5 |
| Frontiermath Tiers 1 3 V2 | 62.8% | #23 / 54 | #37 / 86 | 93.7% GPT-6 Astra |
| ARC-AGI 2 | 72.1% | #12 / 45 (tied with 1) | #19 / 83 | 95.0% GPT-6 Astra |
| ALE-Bench | 911 | #34 / 56 | #47 / 82 | 2,951 GPT-6 Astra |
| WebDev Arena | 1,506 | #27 / 64 | #42 / 82 | 1,820 Claude Opus 5.5 |
| ARC-AGI | 92.5% | #14 / 45 (tied with 1) | #21 / 81 | 98.5% Claude Fable 5 |
| SimpleQA Verified | 66.2% | #6 / 53 (tied with 1) | #12 / 81 | 76.0% Gemini 4 Argon |
| Chess puzzles | 45.0% | #16 / 52 (tied with 2) | #26 / 80 | 76.0% GPT-6 Astra |
| Game puzzles | 32.0% | #16 / 48 (tied with 1) | #26 / 77 | 84.0% GPT-6 Astra |
| ProofBench | 31.0% | #25 / 51 (tied with 1) | #42 / 74 | 100.0% Claude Fable 5.1 |
| SimpleBench | 76.7% | #5 / 41 | #5 / 74 | 81.9% Claude Fable 5 |
| Frontiermath Tier 4 V2 | 26.8% | #22 / 43 (tied with 3) | #34 / 67 | 100.0% GPT-6.1 Sol |
| FrontierMath | 39.0% | #10 / 33 (tied with 1) | #10 / 65 | 51.7% GPT-5.5 |
| ForecastBench | 59.10 | #19 / 25 | #35 / 62 | 62.50 o3 |
| Vending Bench 2 | 5,396 | #20 / 44 | #27 / 57 | 15,515 GPT-6 Astra |
| Apex Agents | 27.5% | #31 / 34 | #46 / 50 | 82.4% Gemini 4 Argon |
| FrontierMath tier 4 | 14.6% | #12 / 33 | #12 / 48 | 39.6% GPT-5.5 Pro |
| Gdp Pdf | 14.0% | #17 / 23 (tied with 2) | #30 / 37 | 35.0% GPT-6.1 Sol |
| EnigmaEval | 25.4% | #4 / 17 | #4 / 36 | 39.3% Claude Fable 5 |
| SWE-bench Verified | 79.3% | #3 / 23 | #3 / 32 | 83.5% Claude Opus 4.7 |
| Blueprint Bench 2 | 33.6% | #3 / 22 (tied with 1) | #9 / 31 | 54.4% Gemini 4 Argon |
| Ebr Bench | 4.8% | #15 / 17 | #25 / 27 | 76.2% GPT-6 Astra |
| DeepSWE | 37.4% | #20 / 23 | #23 / 26 | 74.1% GPT-6 Astra |
| Surface Evolver Bench | 58.1% | #10 / 19 | #11 / 24 | 95.0% Claude Fable 5 |
| Gbaeval | 6.7% | #14 / 22 | #14 / 22 | 79.6% Claude Opus 5 |
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 3.5 Flash 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 |
|---|---|---|---|---|---|
| Muse Spark 1.2 | Meta AI | 154.90 | 1,509 | 78.0 | Aug 5, 2026 |
| DeepSeek V4 Flash 0731 | DeepSeek | 154.49 | n/a | 74.2 | Jul 31, 2026 |
| Gemini 3.5 Flash | Google DeepMind | 154.38 | 1,526 | 74.6 | May 19, 2026 |
| Grok 4.7 | xAI | 154.37 | 1,453 | 77.4 | Sep 21, 2026 |
| Muse Spark 1.1 | Meta AI | 154.00 | 1,516 | 75.3 | Jul 9, 2026 |
| Grok 4.5 | xAI | 153.93 | 1,495 | 75.8 | Jul 8, 2026 |
| GPT-5.2 Codex | OpenAI | no ECI | n/a | 74.0 | Dec 18, 2025 |
Gemini 3.5 Flash or Grok 4.7?
Grok 4.7 is the model from another lab closest to Gemini 3.5 Flash in capability. Across the 19 tests both took, this is the score:
| Test | Gemini 3.5 Flash | Grok 4.7 |
|---|---|---|
| Apex Agents | 27.5% | 54.6% |
| Vending Bench 2 | 5,396 | 10,537 |
| Gdp Pdf | 14.0% | 22.8% |
| Frontiermath Tier 4 V2 | 26.8% | 17.1% |
| CritPt | 13.1% | 18.0% |
| Frontiermath Tiers 1 3 V2 | 62.8% | 53.0% |
| SimpleQA Verified | 66.2% | 56.0% |
| ARC-AGI 2 | 72.1% | 61.4% |
| Game puzzles | 32.0% | 29.0% |
| Chess puzzles | 45.0% | 41.0% |
Unweighted score: each test counts one point, whether math or coding. The dot marks who did better on each row.
What Gemini 3.5 Flash costs to deliver each task
Price per token is what is on the price sheet; cost per task is what lands on the invoice, because it depends on how many tokens the model spends. Each point is an effort level, published by whoever ran the test and compiled by Epoch AI. The further up and to the left, the more score per dollar.
ARC-AGI 2
The score scale does not start at zero.
ARC-AGI
The score scale does not start at zero.
DeepSWE
The score scale does not start at zero.
Other models from Google DeepMind
The models from the same lab closest in time, newest first.
| Model | Release | ECI | ELO | LiveBench | Access |
|---|---|---|---|---|---|
| Gemini 4 Argon | Sep 30, 2026 | 164.56 | 1,608 | n/a | Limited access |
| Gemini 3.8 Flash | Sep 2, 2026 | 157.20 | 1,551 | 75.8 | API access |
| Gemini 3.7 Flash | Aug 13, 2026 | 157.22 | 1,551 | 78.8 | API access |
| Gemini 3.6 Flash | Jul 21, 2026 | 154.24 | 1,530 | 73.6 | API access |
| Gemini 3.5 Flash-Lite | Jul 21, 2026 | 145.15 | 1,467 | 63.9 | API access |
| Gemini 3.5 Flash | May 19, 2026 | 154.38 | 1,526 | 74.6 | API access |
| Gemma 4 31B IT | Apr 2, 2026 | 142.86 | n/a | n/a | Open weights (restricted use) |
| Gemma 4 26B A4B | Apr 2, 2026 | 141.81 | 1,495 | n/a | Open weights (unrestricted) |
| Gemini 3.1 Pro | Feb 19, 2026 | 154.67 | 1,528 | 77.0 | API access |
Have you used Gemini 3.5 Flash? 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 3.5 Flash
Where does Gemini 3.5 Flash rank among AI models?
In the ROO3 AI Benchmark, ordered by the Epoch AI ECI, Gemini 3.5 Flash ranks #37 out of 267 models, with an ECI of 154.38. 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 3.5 Flash better than Grok 4.7?
Across the 19 tests both models took and Epoch AI compiled, Gemini 3.5 Flash comes out ahead on 9, behind on 10 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 3.5 Flash 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 3.5 Flash are: SWE-bench Verified (#3 of 23); SimpleBench (#5 of 41); SimpleQA Verified (#6 of 53). 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