Share your thoughts, 1 month free Claude Pro on usSee more

Laboratory Science Knowledge Evaluation on LAB-Bench 2 821 (Evaluation)

82.3Accuracy

Claude Opus 4.6

Updated 26d ago

Evaluation Results

Method	Links
Claude Opus 4.6 2026.05		82.3
SOTA 2026.05		80
Claude Sonnet 4.6 2026.05		74.3
Claude Opus 4.5 2026.05		73.8
GPT-5.4 2026.05		72.1
Claude Sonnet 4.5 2026.05		71.5
Gemini 3.1 Pro 2026.05		70.9
Nemotron-3 Super 120B 2026.05		62.6
GLM-4.5 2026.05		61.5
Trinity-Large 2026.05		61
Gemini 3 Flash 2026.05		59.8
INTELLECT-3.1 2026.05		57.9
Qwen3-235B 2026.05		57.9
Gemini 3.1 Pro 2026.05		52.1
Gemini 3 Flash 2026.05		51.9
Claude Opus 4.6 2026.05		51.2
Claude Opus 4.5 2026.05		49.8
Claude Sonnet 4.6 2026.05		49.3
GPT-5.4 2026.05		48.5
Claude Sonnet 4.5 2026.05		48.1
GLM-4.5 2026.05		44.8
Nemotron-3 Super 120B 2026.05		43.6
Trinity-Large 2026.05		43.5
INTELLECT-3.1 2026.05		43.5
Qwen3-235B 2026.05		42.5