whatmodel.sh / Gemini 3.7 Flash

Gemini 3.7 Flash

+ contribute
Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Gemini 3.7 Flash is a closed-weights language model from Google, released August 2026.

76.3±0.6#32 of 719 LLMsfrom 31 benchmark results across 9 sources

$ how it compares

The LLMs ranked either side of Gemini 3.7 Flash, out of 828 with a score, then more from Google.

#modelscorewhereRAMreleased
30Grok 4.6 SpaceXAI (xAI)77±1api $3.00/M2026-08
31DeepSeek V4 Pro (0813) DeepSeek76±1api $1.63/M +local~1010GB2026-08
32Gemini 3.7 Flash Google76±1api $1.50/M2026-08
33GLM-5.3 Z.ai76±1api $2.15/M +local~239GB2026-08
34Qwen3.8 Max (0902) Alibaba76±1api $3.00/M2026-09
more from Google
9Gemini 4 Argon (limited preview) Google82±1limited preview2026-09
23Gemini 3.8 Flash Google77±1api $1.50/M2026-09
39Gemini 3.1 Pro (Preview) Google75±1api $4.50/M2026-02

see all LLMs · find the ones that fit your machine

$ what it's good at

agent tasks▁▁▃▆█▆▅▅▅▄▃▁ #41 of 300 · 18% of its score · 8 tests

Trails GLM-5.2, GLM-5.3-Flash, Grok 4.7 here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
Arena Agent (October 2026 aggregate IPS)-0.0149 IPS ✎▁▁▁▄▂▃▇█▂▂▁▃ beats 15/49avg -0.312 IPS▇▇ 6.0%sharp
Terminal-Bench 2.1 lab +281.6% ✎▁·▁·▁▁▄▃▄█▃▇ beats 49/56hard · avg 23%▇ 3.6%small run · several runs
Terminal-Bench 4.011.2% ✎█▅▅▅▅·▅▃▃▅▅▅ beats 0/20very hard · avg 0.4%▇ 3.3%sharp
SWE-bench Pro (split unspecified) lab60.4% ✎▃·▁▁▁▂▂▃█▇▃▁ beats 41/55very hard · avg 6.4%▇ 2.0%sharp · at its level · small run
DeepSWE v1.1 (resolved) lab65.3% ✎▂····▁▁▁▆██▆ beats 11/28hard · avg 23%▇ 1.0%at its level · small run
KlinikeBench strict pass@1 (333 interactive clinical encounters, Terminus-2)10.8% ✎▁▁·▁▁▁▁▅▃█▅█ beats 11/27very hard · avg 2.4%▇ 0.9%typical
Terminal-Bench 3.0 lab14.9% ✎█·█··██····█ beats 3/4very hard · avg 1.7%▇ 0.7%small run
τ³-bench Banking lab30.9% ✎▂▂·▅█▂▂▃▂▅▂▃ beats 19/20very hard · avg 6.9%▇ 0.5%blunt · small run
coding▁▁▁▃▆██▇▆▄▃▁ #27 of 458 · 10% of its score · 4 tests

Beats GPT-5.5, GPT-5.6 Terra, Claude Opus 4.8 here, though they rank higher overall. Trails GLM-5.2, GLM-5.3-Flash, Grok 4.7 here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
Arena WebDev (Code Arena)1591.7 ✎▁▂▄▃▅█▄▇▆▄▁▁ beats 95/119avg 981▇▇ 5.3%sharp
SWE-bench Pro (split unspecified) lab60.4% ✎▃·▁▁▁▂▂▃█▇▃▁ beats 41/55very hard · avg 6.4%▇ 2.0%sharp · at its level · small run
Luau Leetcode fully solved (Torpedo Software)79.8% ✎▁▁▁▁·▁▂▃▅▃█▇ beats 32/49hard · avg 18%▇ 1.9%typical
DeepSWE v1.1 (resolved) lab65.3% ✎▂····▁▁▁▆██▆ beats 11/28hard · avg 23%▇ 1.0%at its level · small run
chat & writing▁▁▁▃▅▄▅█▆▆▇▃ #13 of 393 * · 3.0% of its score · 1 test

Beats GPT-6 Astra, GPT-6.1 Sol, Claude Sonnet 5.5 here, though they rank higher overall. Trails Claude Opus 4.6, Muse Spark 1.1, Claude Opus 4.7 here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
LMArena (text)1487.7 ✎▁▁▁▂▃▃▃▆▆▇█▃ beats 366/379avg 1268▇ 3.0%blunt
math▁▂▃▅▆▅▄█▄▃▁▁ #24 of 398 · 48% of its score · 8 tests

Beats DeepSeek V4.1 Flash, Gemini 3.8 Flash, Grok 4.6 here, though they rank higher overall.

testresultvs the others in its rundifficultycountswhy
FrontierMath tier 4 v236.6% ✎▁▂▁▂█▅▂▂▁▁▁▁ beats 41/63very hard · avg 0.1%▇▇▇▇▇ 13%sharp · at its level
FrontierMath tiers 1–3 v271.6% ✎▂··▁▁▃▄▄█▅▃▃ beats 60/82very hard · avg 1.2%▇▇▇▇▇ 13%sharp · at its level
MathArena ArXivMath June 2026 (48 problems)62.5% ✎▂···▄·██▆▆▄▂ beats 7/19very hard · avg 4.6%▇▇▇ 7.5%sharp · at its level
MathArena ArXivMath April 2026 (40 problems)57.5% ✎▁·····▅██▄·▁ beats 11/20very hard · avg 7.2%▇▇▇ 6.6%sharp · at its level
MathArena BrokenArXiv April 2026 (61 problems)19.3% ✎▂▂▆█▃▃▂▂▂··▂ beats 6/18very hard · avg 0.7%▇ 2.8%sharp · at its level · noisy
MathArena ArXivMath May 2026 (40 problems)47.5% ✎▁··▁▄█▄▁·▃·▁ beats 4/17very hard · avg 10%▇ 2.3%at its level
MathArena ArXivLean June 2026 (46 problems)21.7% ✎▄····▄▄███·▄ beats 5/9very hard · avg 0.7%▇ 2.0%sharp
OTIS mock AIME 2024–202597.2% ✎▃▃▄▃▃▃▇▇█▃▃▅ beats 170/193very hard · avg 8.5%▇ 0.5%sharp · maxed out · noisy
vision & documents▁▁▃▂▁▃▄▂▆▇█▄ #5 of 147 * · 4.9% of its score · 2 tests

Beats GPT-6 Astra, GPT-6.1 Sol, Claude Fable 5.1 here, though they rank higher overall. Trails Claude Opus 4.7 here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
CharXiv (reasoning questions) lab84.5% ✎▃·▃·▃▃▃·▃▃█▅ beats 9/11medium · avg 40%▇ 2.8%typical
Arena vision (style controlled)1296.0 ✎▁▁▂▂▂▂▄▅▅▆██ beats 141/145avg 1068▇ 2.1%blunt
other tests9 not in a task · 19% of its score
testresultvs the others in its rundifficultycountswhy
Furniture Assembly (Epoch)26.7% ✎█▆▆▆▆▂▅▂▂▂·▅ beats 6/30very hard · avg 1.2%▇▇ 5.2%sharp · at its level
Mystery Game Puzzles (Epoch)37.0% ✎█··▂▃▇▄█▂▃▁▁ beats 59/74very hard · avg 9.4%▇▇ 4.4%sharp · noisy
Chess Puzzles (Epoch)47.0% ✎█▁▁▁▁▂▁▄▂▃▁▁ beats 124/135very hard · avg 5.9%▇ 2.7%sharp · at its level · noisy
GPQA Diamond94.8% ✎▁▁▁▁▂▄▆▄▅▇█▅ beats 204/208hard · avg 40%▇ 2.0%sharp
SimpleQA Verified69.2% ✎▁▁▃▂▃▃▆▆█▃▂▅ beats 70/78very hard · avg 8.0%▇ 1.2%at its level
Icelandic LLM Leaderboard average (Miðeind, Sept 2026 suite)65.0% ✎▁▁▃▂▂▄▁▃▄▄▄█ beats 30/38very hard · avg 14%▇ 1.1%at its level
MRCR v2 8-needle, 128k lab97.0% ✎█▄▄█·▄···▄·▄ beats 8/8medium · avg 47%▇ 1.0%small run
LVBench lab85.4% ✎▇·█▅▅▄▄··▃▁▁ beats 26/28medium · avg 57%▇ 1.0%typical
1 more test
testresultvs the others in its rundifficultycountswhy
RobloxQA v2.0 multiple choice (Torpedo Software)96.8% ✎▁▁▃▁█▄▃▃▆▂▂▂ beats 40/49easy · avg 78%▇ 0.6%maxed out

$ where to run it

Through an API

api · $1.50/M · $0.75 in / $3.75 out per 1M tokens · 1M context · ai.google.dev ↗ · price sourceStandard base token rate from models.dev. Provider-specific tiers, long-context surcharges, caching and promotions may differ.

$ how the score was worked out · Epoch AI — internal evaluatio… 42%, MathArena ArXivMath / BrokenA… 21%, LMArena leaderboard dataset 16% and 6 more

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

sourceresultsshare of the scorelicence
Epoch AI — internal evaluations8▇▇▇▇▇▇▇▇▇▇▇▇▇ 42%CC BY 4.0
MathArena ArXivMath / BrokenArXiv / ArXivLean model-output datasets5▇▇▇▇▇▇ 21%CC BY-SA 4.0
LMArena leaderboard dataset4▇▇▇▇▇ 16%CC BY 4.0
Google DeepMind Gemini model page and model cards (lab-reported scores) · lab-reported6▇▇▇ 8%Factual vendor-reported score (not licensed content)
Google Cloud developer's guides to Gemini 3.8 Flash and 3.7 Flash · lab-reported4▇ 4%CC-BY-4.0
Terminal-Bench 4.0 official leaderboard submissions1▇ 3%Apache-2.0
Roblox LLM Leaderboard (Torpedo Software): RobloxQA v2.0 and Luau Leetcode2▇ 3%MIT
Miðeind Icelandic LLM Leaderboard1▇ 1%CC BY 4.0
KlinikeBench: Evaluating Language Models Beyond Diagnostic Accuracy (arXiv:2609.38480v2)1▇ 1%CC BY 4.0
prior · pull toward the middle▇ 0%
weighted average= 76.3

$ improve this page

Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

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