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-opus-5.5

Published benchmark results for this specific model configuration.

claude-opus-5.5
Overall score84.7 / 100
Cost / successful task$6.622USD · benchmark workload
Weight accessNot reported

Where this configuration performs

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

Long Agentic72.8
Reasoning92.2
Coding87.1
Browser88.4
Data Analysis80.3
Research88.9
Design83.3

Task-level results

long agentic72.8reasoning92.2coding87.1browser88.4data analysis80.3research88.9design83.3

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