Gemini 3.7 Flash: benchmarks, price and rank
Gemini 3.7 Flash from Google is 17th of 123 rated AI models on Spezia's leaderboard, strongest on MRCR v2 (8 needles) and GPQA Diamond. It costs $0.75 in, $3.75 out per million tokens.
- 17th of 123 rated models
- Rating 1,661
- Rated on 13 benchmarks
- Closed
- Released Aug 13, 2026
- 1 million tokens of context
- Knows the world to March 2026
Where it is strongest
- 2nd of 24 on MRCR v2 (8 needles)
- 7th of 103 on GPQA Diamond
Where it falls behind
- 17th of 21 on FrontierSWE
- 19th of 27 on DeepSWE 1.1
- 10th of 17 on Remote Labor Index
Price: $0.75 in, $3.75 out per million tokens.
Its headline results
The benchmarks that count most in the rating, with its place among the models measured the same way, and where each number came from.
- Remote Labor Index5%
Real freelance projects that people were paid to do, from design and video to architecture, games and data work. The model gets the same brief and files the freelancer got and has to hand back finished work.
10th of 17. Independent result from Epoch AI.
- APEX-Agents67.8%
Long tasks written by investment banking analysts, management consultants and corporate lawyers from their own work, done across files and apps in a realistic work setup.
9th of 56. Independent result from Epoch AI.
- FrontierSWE20.3%
Open-ended technical projects that take hours to tens of hours, from making systems faster to building large programs and applied machine learning research. Each run gets a 20-hour budget.
17th of 21. Independent result from Epoch AI.
- DeepSWE 1.165.3%
Software engineering tasks run in the same simple agent setup for every model.
19th of 27. The maker's own number, from Google, via LLM Stats.
- SimpleQA Verified69.2%
Short factual questions with one checked answer, asked without web search.
12th of 78. Independent result from Epoch AI.
- FrontierMath71.6%
Hundreds of new, unpublished math problems written by mathematicians, from hard undergraduate level up to research, with answers a computer can check.
29th of 76. Independent result from Epoch AI.
- CharXiv Reasoning88.7%
Questions about real charts from scientific papers that need reasoning across the chart, not just reading off a value.
10th of 45. The maker's own number, from Google, via LLM Stats.
- MRCR v2 (8 needles)97%
A very long conversation full of similar requests, where the model has to find and repeat particular earlier answers, eight of them at once.
2nd of 24. The maker's own number, from Google, via LLM Stats.
- LMArena text1,489
People put a question to two anonymous models and vote for the better answer.
12th of 107. Independent result from LMArena.
- LMArena WebDev1,588
People ask two anonymous models to build the same web page or app and vote for the better one.
30th of 98. Independent result from LMArena.
5 more results in its full profile.
Every result with its source, what it costs per task, what jumped with its release, side by side with any other model, and answers to your questions about when to use it.
Models ranked near it
How the rating works, and every benchmark it uses: the model leaderboard.