Google · LIVEBENCH 2026-06-25

Gemini 3.7 Flash High

Published benchmark results for this specific model configuration.

gemini-3.7-flash-high

Overall score78.8 / 100

Cost / successful task$0.157USD · benchmark workload

Weight accessNot reported

Where this configuration performs

Category averages from the same release, on a 0–100 scale.

Reasoning87.8

Coding78.9

Agentic Coding58.3

Mathematics93.5

Data Analysis68.0

Language85.5

Instruction Following79.9

Task-level results

theory of mind82.7zebra puzzle94.5spatial100.0logic with navigation74.0code generation81.7code completion76.1javascript68.2typescript46.7python60.0AMPS Hard98.0integrals with game88.0math comp96.1olympiad91.8consecutive events60.4tablejoin45.5tablereformat98.0connections100.0plot unscrambling66.4typos90.0paraphrase78.1simplify74.6story generation82.3summarize84.8

Source files: Scores ↗ Categories ↗ Costs ↗ . Missing cost is unknown, never free. Compare configurations within the same release.