BenchLM comparison
Claude Opus 5.5
Anthropic
86.66 /100
Supported · public rank #1
vs
GPT-6.1 Sol
OpenAI
84.21 /100
Supported · public rank #2
CategoryClaude Opus 5.5GPT-6.1 SolReading
Agentic90.4
Supported · #1/130
76.2
Estimated · #5/130
Directional only
Coding82.7
Supported · #2/151
71
Supported · #8/151
Claude Opus 5.5 leads
Reasoning86.1
#4/27
90.5
#2/27
Directional only
Multimodal92.2
#3/55
87.3
Unranked · 1 rankable row
Not comparable
Knowledge90.2
Supported · #1/186
86.1
Supported · #4/186
Claude Opus 5.5 leads
Multilingual——Not comparable
Instruction following——Not comparable
Math97.3
Unranked · 4 rankable rows
97.3
Unranked · 2 rankable rows
Not comparable
Methodology bench-align-v5.8-2026-09-30Claude Opus 5.5 vs GPT-6.1 Sol benchmark comparison by BenchLM.ai