aitoolcompass.

Find your next AI tool

Search by product name or task. Press Escape to close.

← All model benchmarks

Unspecified · LIVEBENCH 2026-11-06

gemma-4-31b

Published benchmark results for this specific model configuration.

gemma-4-31b
Overall score35.0 / 100
Cost / successful task$0.202USD · benchmark workload
Weight accessNot reported

Where this configuration performs

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

Long Agentic8.5
Reasoning68.3
Coding20.3
Browser0.0
Data Analysis53.3
Research38.1
Design56.5

Task-level results

long agentic8.5reasoning68.3coding20.3browser0.0data analysis53.3research38.1design56.5

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