BenchLM comparison
Claude Opus 5.5
Anthropic
86.29 /100
Supported · public rank #2
vs
Gemini 4 Carbon
Google
88.38 /100
Estimated · public rank #1
CategoryClaude Opus 5.5Gemini 4 CarbonReading
Agentic90.1
Supported · #1/135
—Not comparable
Coding82.9
Supported · #1/156
81.1
Supported · #3/156
Claude Opus 5.5 leads
Reasoning86.1
#4/26
79.5
Unranked · 2 rankable rows
Not comparable
Multimodal91.8
#3/58
—Not comparable
Knowledge89.5
Supported · #1/179
88.2
Supported · #2/179
Claude Opus 5.5 leads
Multilingual——Not comparable
Instruction following——Not comparable
Math97.3
Unranked · 4 rankable rows
—Not comparable
Methodology bench-align-v5.8-2026-09-30Claude Opus 5.5 vs Gemini 4 Carbon benchmark comparison by BenchLM.ai