whatmodel.sh / Gemini 3.5 Flash

Gemini 3.5 Flash

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

Gemini 3.5 Flash is a closed-weights language model from Google, released May 2026.

74.0±0.6#50 of 719 LLMsfrom 30 benchmark results across 9 sources

$ how it compares

The LLMs ranked either side of Gemini 3.5 Flash, out of 828 with a score, then more from Google.

#modelscorewhereRAMreleased
51Qwen3.7 Max Alibaba74±1api $3.75/M2026-05
52GLM-5.3-Flash Z.ai74±1api $0.24/M +local~103GB2026-08
53Gemini 3.5 Flash Google74±1api $3.38/M2026-05
54GPT-5.2 OpenAI74±1api $4.81/M2025-12
55Hunyuan Hy3 (295B-A21B) Tencent74±1api $0.23/M +local~96GB2026-07
more from Google
9Gemini 4 Argon (limited preview) Google82±1limited preview2026-09
23Gemini 3.8 Flash Google77±1api $1.50/M2026-09
32Gemini 3.7 Flash Google76±1api $1.50/M2026-08

see all LLMs · find the ones that fit your machine

$ what it's good at

agent tasks▁▁▃▆█▆▅▅▅▄▃▁ #49 of 300 · 8.0% of its score · 4 tests

Beats Gemini 3.1 Pro (Preview), Muse Spark 1.2, Muse Spark 1.1 here, though they rank higher overall. Trails Claude Sonnet 4.6, Qwen3.8 27B, GLM-5.2 here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
SWE-bench Verified79.3% ✎▁··▁·▁▂▁▄█▄▁ beats 28/29hard · avg 19%▇▇ 5.2%sharp
Legal Research Bench (agentic, all-pass)30.8% ✎▄█··▄▄▄██·▄█ beats 8/12very hard · avg 5.8%▇ 1.1%at its level
EEGAgentBench overall (tool-based EEG analysis agents, 6 tasks)49.5% ✎▂·▁▁··▇▇█▂·▁ beats 18/25hard · avg 29%▇ 0.9%blunt · at its level
KlinikeBench strict pass@1 (333 interactive clinical encounters, Terminus-2)15.6% ✎▁▁·▁▁▁▁▅▃█▅█ beats 14/27very hard · avg 2.4%▇ 0.8%typical
coding▁▁▁▃▆██▇▆▄▃▁ #52 of 458 * · 11% of its score · 2 tests

Beats GPT-5.4, DeepSeek V4 Flash Vision (Exp), Gemini 3.1 Pro (Preview) here, though they rank higher overall. Trails Mistral Large 4 (public preview), Qwen3.7 Plus, Claude Sonnet 4.6 here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
Arena WebDev (Code Arena)1499.0 ✎▁▂▄▃▅█▄▇▆▄▁▁ beats 69/119avg 981▇▇ 5.4%sharp
SWE-bench Verified79.3% ✎▁··▁·▁▂▁▄█▄▁ beats 28/29hard · avg 19%▇▇ 5.2%sharp
chat & writing▁▁▁▃▅▄▅█▆▆▇▃ #24 of 393 * · 3.0% of its score · 1 test

Beats Claude Sonnet 5.5, GPT-6 Sol, GPT-5.6 Terra here, though they rank higher overall. Trails Gemini 3.6 Flash here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
LMArena (text)1477.4 ✎▁▁▁▂▃▃▃▆▆▇█▃ beats 355/379avg 1268▇ 3.0%blunt
math▁▂▃▅▆▅▄█▄▁▁▁ #40 of 398 · 60% of its score · 13 tests

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

testresultvs the others in its rundifficultycountswhy
FrontierMath tiers 1–3 v262.8% ✎▂··▁▁▃▄▄█▅▃▃ beats 48/82very hard · avg 1.2%▇▇▇▇▇▇ 15%sharp · at its level
FrontierMath tier 4 v226.8% ✎▁▂▁▂█▅▂▂▁▁▁▁ beats 29/63very hard · avg 0.1%▇▇▇▇ 11%sharp · at its level
MathArena ArXivMath June 2026 (48 problems)52.1% ✎▂···▄·██▆▆▄▂ beats 4/19very hard · avg 4.6%▇▇▇ 8.2%sharp · at its level
MathArena ArXivMath April 2026 (40 problems)52.5% ✎▁·····▅██▄·▁ beats 9/20very hard · avg 7.2%▇▇▇ 6.9%sharp · at its level
MathArena ArXivMath February 2026 (32 problems)50.8% ✎▁▁▁▁█▇▄▃▁▃·▃ beats 18/24very hard · avg 9.1%▇▇ 3.9%at its level
MathArena ArXivMath March 2026 (30 problems)55.8% ✎█▃▅·█▃▃·▅·▃▃ beats 8/14very hard · avg 15%▇ 3.4%at its level
MathArena BrokenArXiv April 2026 (61 problems)21.3% ✎▂▂▆█▃▃▂▂▂··▂ beats 7/18very hard · avg 0.7%▇ 2.5%sharp · at its level · noisy
MathArena ArXivMath May 2026 (40 problems)44.2% ✎▁··▁▄█▄▁·▃·▁ beats 2/17very hard · avg 10%▇ 2.5%at its level
5 more tests
testresultvs the others in its rundifficultycountswhy
MathArena BrokenArXiv June 2026 (54 problems)12.3% ✎▄··█▂▂▂▆▄··▂ beats 3/14very hard · avg 1.1%▇ 1.8%sharp · at its level · noisy
MathArena BrokenArXiv May 2026 (50 problems)11.0% ✎██·▄▄▄····▄▄ beats 3/8very hard · avg 1.6%▇ 1.6%sharp
MathArena BrokenArXiv March 2026 (56 problems)16.1% ✎█·▆█▂·▂▂···▂ beats 10/14very hard · avg 0.8%▇ 1.5%sharp · noisy
MathArena BrokenArXiv February 2026 (31 problems)6.5% ✎▃▂▃▃█▃▂·▃··▂ beats 3/16very hard · avg 0.7%▇ 1.4%sharp · noisy
OTIS mock AIME 2024–202595.6% ✎▃▃▄▃▃▃▇▇█▃▂▅ beats 163/193very hard · avg 8.5%▇ 0.7%sharp · maxed out · noisy
vision & documents▁▁▃▂▁▃▄▂▆▇█▄ #16 of 147 · 6.7% of its score · 4 tests

Beats Claude Sonnet 5.5, GPT-5.5, MiMo-V2.6-Pro here, though they rank higher overall.

testresultvs the others in its rundifficultycountswhy
CharXiv (reasoning questions) lab84.2% ✎▃·▃·▃▃▃·▃▃█▅ beats 8/11medium · avg 40%▇ 3.3%typical
Arena vision (style controlled)1284.1 ✎▁▁▂▂▂▂▄▅▅▆██ beats 129/145avg 1068▇ 2.1%blunt
Arena document (style controlled)1453.4 ✎▁▁▃▃▄▃▇█▅▇▄▅ beats 15/37avg 1335▇ 1.0%blunt
SpaceCast-Bench predictive spatial reasoning (overall, 16 task types)58.0% ✎▃·▅▃▅▃··▅··█ beats 11/11very hard · avg 32%▇ 0.3%blunt · at its level · noisy
web search▂▁▂▆▄█▃▃▁▁·▁ #42 of 80 * · 1.1% of its score · 1 test

Beats Gemini 3.1 Pro (Preview), Qwen3.7 Max here, though they rank higher overall. Trails GPT-4o mini (2024-07-18), Mistral Small 3.1 24B (2503), Llama 4 Scout (109B-A17B) here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
Legal Research Bench (agentic, all-pass)30.8% ✎▄█··▄▄▄██·▄█ beats 8/12very hard · avg 5.8%▇ 1.1%at its level
other tests7 not in a task · 17% of its score
testresultvs the others in its rundifficultycountswhy
EBR-bench (Epoch)4.8% ✎▄▃·▆█▂▆▄▆▆·▄ beats 2/23very hard · avg 0.3%▇▇ 5.0%sharp
Mystery Game Puzzles (Epoch)32.0% ✎█··▂▃▇▄█▂▃▁▁ beats 50/74very hard · avg 9.4%▇ 3.4%sharp · noisy
GPQA Diamond92.8% ✎▁▁▁▁▂▄▆▄▅▇█▅ beats 192/208hard · avg 40%▇ 2.6%sharp
Chess Puzzles (Epoch)50.0% ✎█▁▁▁▁▂▁▄▂▂▃▁ beats 129/135very hard · avg 5.9%▇ 2.3%sharp · noisy
Legal Research Bench (single-shot, no tools, all-pass)6.3% ✎▃▃█▃·▅···█▃▃ beats 8/12very hard · avg 0.8%▇ 1.3%too hard for it
SimpleQA Verified66.2% ✎▁▁▃▂▃▃▆▆█▃▃▅ beats 68/78very hard · avg 8.0%▇ 1.2%at its level
Icelandic LLM Leaderboard average (Miðeind, Sept 2026 suite)62.2% ✎▁▁▃▂▂▄▁▃▄▄▄█ beats 27/38very hard · avg 14%▇ 1.1%at its level

$ where to run it

Through an API

api · $3.38/M · $1.5 in / $9 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… 46%, MathArena ArXivMath / BrokenA… 34%, LMArena leaderboard dataset 11% 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 evaluations9▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 46%CC BY 4.0
MathArena ArXivMath / BrokenArXiv / ArXivLean model-output datasets10▇▇▇▇▇▇▇▇▇▇ 34%CC BY-SA 4.0
LMArena leaderboard dataset4▇▇▇ 11%CC BY 4.0
Google DeepMind Gemini model page and model cards (lab-reported scores) · lab-reported1▇ 3%Factual vendor-reported score (not licensed content)
Miðeind Icelandic LLM Leaderboard1▇ 1%CC BY 4.0
EEGAgentBench: Benchmarking LLM Agents on Short- and Long-Horizon EEG Analysis (arXiv:2609.31632v1)1▇ 1%CC BY 4.0
KlinikeBench: Evaluating Language Models Beyond Diagnostic Accuracy (arXiv:2609.38480v2)1▇ 1%CC BY 4.0
SpaceCast-Bench: Evaluating Predictive Spatial Reasoning in Vision-Language Models (arXiv:2610.12402v1)1▇ 0%CC BY 4.0
prior · pull toward the middle▇ 0%
weighted average= 74.0

$ 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