aitoolcompass.

Find your next AI tool

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

← All model benchmarks

Google · LIVEBENCH 2026-11-06

gemini-4-argon

Published benchmark results for this specific model configuration.

gemini-4-argon
Overall score80.5 / 100
Cost / successful task$1.759USD · benchmark workload
Weight accessNot reported

Where this configuration performs

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

Long Agentic58.5
Reasoning95.1
Coding85.7
Browser82.8
Data Analysis85.2
Research86.8
Design69.2

Task-level results

long agentic58.5reasoning95.1coding85.7browser82.8data analysis85.2research86.8design69.2

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