Google · LIVEBENCH 2026-06-25

Gemini 3.1 Pro Preview High

Published benchmark results for this specific model configuration.

gemini-3.1-pro-preview-high

Overall score77.0 / 100

Cost / successful task$0.286USD · benchmark workload

Weight accessNot reported

Where this configuration performs

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

Reasoning84.0

Coding76.5

Agentic Coding44.1

Mathematics91.0

Data Analysis78.5

Language85.4

Instruction Following79.1

Task-level results

theory of mind80.8zebra puzzle85.3spatial98.0logic with navigation72.0code generation74.6code completion78.3javascript59.1typescript23.3python50.0AMPS Hard98.0integrals with game78.0math comp96.1olympiad92.1consecutive events85.2tablejoin52.3tablereformat98.0connections100.0plot unscrambling74.1typos82.0paraphrase79.9simplify71.0story generation77.3summarize88.2

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