aitoolcompass.

Find your next AI tool

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

← All model benchmarks

Anthropic · LIVEBENCH 2026-11-06

claude-sonnet-5.5

Published benchmark results for this specific model configuration.

claude-sonnet-5.5
Overall score78.6 / 100
Cost / successful task$5.683USD · benchmark workload
Weight accessNot reported

Where this configuration performs

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

Long Agentic63.4
Reasoning91.6
Coding77.2
Browser89.3
Data Analysis59.5
Research88.0
Design81.3

Task-level results

long agentic63.4reasoning91.6coding77.2browser89.3data analysis59.5research88.0design81.3

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