whatmodel.sh / Gemini 1.5 Flash Experimental (0827)
Gemini 1.5 Flash Experimental (0827)
↗ share
+ contribute
Gemini 1.5 Flash Experimental (0827) is a closed-weights language model from Google, released August 2024. Licence: proprietary.
Catalogue identity notes
Comparison-only: retired or no longer listed API model, kept so its independent runs can be scored. Gemini API changelog, August 27, 2024: experimental gemini-1.5-flash-exp-0827 released; shut down. Distinct from the stable gemini-1.5-flash-002.
$ how it compares
The LLMs ranked either side of Gemini 1.5 Flash Experimental (0827), out of 828 with a score, then more from Google.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 350 |  Mistral Small 3 24B (2501) | 53±1 | local | ||
| 351 |  Qwen2-VL 72B Instruct | 53±1 | local | ||
| 352 | Gemini 1.5 Flash Experimental (0827) | 53±1 | comparison only | ||
| 353 |  GPT-4o mini (2024-07-18) | 53±0 | api $0.26/M | ||
| 354 | Gemini 1.5 Pro (001) | 53±1 | comparison only | ||
| more from Google | |||||
| 9 | Gemini 4 Argon (limited preview) | 82±1 | limited preview | ||
| 23 | Gemini 3.8 Flash | 77±1 | api $1.50/M | ||
| 32 | Gemini 3.7 Flash | 76±1 | api $1.50/M | ||
$ what it's good at
coding▁▁▁▃▅▇█▇▅▄▃▁ #294 of 524 * · 27% of its score · 2 tests
Beats Qwen3-VL 32B Instruct, Solar Open 100B (102B-A12B), GLM-4.7-Flash here, though they rank higher overall. Trails QwQ-32B-Preview, InternLM2.5 20B Chat, Mistral Small Instruct 22B (2409) here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| CRUXEval (ZeroEval, 800 problems) | 74.5% ✎ | ▁▂▆▃▆█▆▃▂·▁▁ beats 34/47 | 15% |
| Aider code editing (133 Exercism Python exercises, two attempts) | 52.6% ✎ | ▁·▁▂▄▁▆▄█▆▂▂ beats 18/62 | 11% |
math▁▂▃▅▇▅▅█▄▁▁▁ #270 of 462 * · 9.4% of its score · 1 test
Beats GPT-4o (2024-08-06), Command A (03-2025), GPT-4o (2024-11-20) here, though they rank higher overall. Trails OpenThinker 7B, Olmo 3 7B Instruct, Tulu 2 70B here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| MATH Level 5 (ZeroEval, 721 problems) | 54.5% ✎ | ▁▁▁▆█▂▇▇▄··▂ beats 29/38 | 9.4% |
other tests3 not in a task · 64% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| LiveBench 2024-07-26 | 46.9% ✎ | ▁▁▁▂▂▃▂█▆▃▅▃ beats 49/61 | 28% |
| MMLU-Redux (ZeroEval, 2,778 questions) | 82.1% ✎ | ▁·▁▁▄▄▄▅█▄▂▁ beats 35/53 | 24% |
| ZebraLogic (ZeroEval, 1,000 puzzles) | 25.0% ✎ | ▁▁▃▆██▁▁·▁▁▁ beats 37/57 | 12% |
$ where to run it
No verified API or download recorded yet · add one
$ how the score was worked out · WildEval ZeroEval published r… 61%, LiveBench model judgments (HF… 28%, Aider code editing benchmark… 11%
Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works
| source | results | share of the score |
|---|---|---|
 WildEval ZeroEval published results | 4 | 61% |
| LiveBench model judgments (HF livebench/model_judgment, 2024-10-22 revision) with the LiveBench question datasets | 1 | 28% |
 Aider code editing benchmark runs (old edit leaderboard) | 1 | 11% |
| prior · pull toward the middle | 0% | |
| weighted average | = 53.2 |
$ improve this page
Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.
data updated · sources and licences · json