Google DeepMind·released 2024-09-241 source
Gemini 1.5 Flash (Sep 2024) benchmark scores: 10 benchmarks tracked. 40% of its scores are independently reproduced — the rest are self-reported or unverified.
Leads on 0 of 10 benchmarks·best result 76 on GeoBench (unverified)·4 of 10 independently reproduced
| Benchmark | Score | Evidence status | Measured |
|---|---|---|---|
| GeoBench | 76accuracy (%) | unverifiedT1 | 2024-09-24GeoBench leaderboard |
| PIQA | 75accuracy (%) | self-reported· optimizedT1 | 2024-09-24Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone |
| MMLU | 65.2accuracy (%) | unverified· optimizedT1 | 2024-09-24Stanford CRFM Leaderboard |
| MATH level 5 | 61.87accuracy (%) | reproduced· optimizedT1 | 2024-09-24Epoch AI |
| GPQA diamond | 29.76accuracy (%) | reproduced· optimizedT1 | 2024-09-24Epoch AI |
| WeirdML | 24.87accuracy (%) | unverifiedT1 | 2024-09-24https://htihle.github.io/weirdml.html |
| DTBench | 22.98accuracy (%) | unverifiedT1 | 2024-09-24https://conceptualreasoning.ai/dtbench |
| OTIS Mock AIME 2024-2025 | 16.17accuracy (%) | reproduced· optimizedT1 | 2024-09-24Epoch AI |
| Balrog | 14.6accuracy (%) | unverifiedT1 | 2024-09-24https://balrogai.com/ |
| FrontierMath-2025-02-28-Private | 0accuracy (%) | reproduced· optimizedT1 | 2024-09-24Epoch AI |
Claims drawn from cited facts, not live model generation.
This model was developed by Google DeepMind and released in 2024. That 2024 vintage places it in a recent generation of work from its developer, though no parameter count is available here to quantify its scale.
4 cited facts
This model is scored on 10 tracked benchmarks, but it holds no current top score on any of them. Its record can be trusted: the results have been independently reproduced, meaning outside evaluation backs the scores up rather than leaving them as unconfirmed vendor claims.
3 cited facts
On average this model trails the leading score by 47.4 points across its benchmarks, a wide score gap under a disclosed harness that reads as well behind the pack rather than a claim about capability. Consistent with that, it holds the top score on 0 benchmarks, meaning there is no category in which it currently leads.
2 cited facts
1 facts cross-checked across data sources: 1 single-source.
Source facts, citations, and refresh stamp for this record.
Sources: epoch_benchmarks
Gemini 1.5 Flash (Sep 2024) is an AI model developed by Google DeepMind, released 2024-09-24. Its benchmark record below tags every score as independently reproduced, self-reported, or unverified.
Gemini 1.5 Flash (Sep 2024) has recorded scores on 10 benchmarks, each shown with its evidence status.
Gemini 1.5 Flash (Sep 2024) has recorded scores on 10 benchmarks — PIQA, GPQA diamond, MATH level 5, OTIS Mock AIME 2024-2025, DTBench, WeirdML, and 4 more. The full table above shows each score with its evidence status.
4 of 10 recorded scores (40%) are independently reproduced rather than self-reported by the lab.
Gemini 1.5 Flash (Sep 2024) has 10 tracked claims: 4 independently reproduced, 1 self-reported, 5 unverified.