Claude Opus 4.5 benchmark ranking on every AI leaderboard

Name: Claude Opus 4.5
Author: Anthropic

closed weights · #23 overall, podium score 6.7 · #1 on 2 boards · top 10 on 2 of 14 · released Nov 24, 2025

14 of 14 boards

Rank	Board	Category	Score	Of
1	BFCLAgents and tools · as “Claude-Opus-4-5-20251101 (FC)”	Agents and tools	77.5%	81
12	MultiChallengeChat · as “claude-opus-4-5-20251101-thinking”	Chat	59.0%	29
14	HLEReasoning · as “claude-opus-4-5-20251101-thinking”	Reasoning	25.2%	42
17	Document ArenaVision · as “claude-opus-4-5-20251101”	Vision	1,462	38
19	Search ArenaAgents and tools · as “claude-opus-4-5-search”	Agents and tools	1,180	33
22	MCP AtlasAgents and tools · as “claude-opus-4-5 (high)”	Agents and tools	69.8%	34
30	Text ArenaChat · as “claude-opus-4-5-20251101-high-32k”	Chat	1,473	100
33	ARC-AGI-2Reasoning · as “Opus 4.5 (Thinking, 64K)”	Reasoning	37.6%	78
42	WebDev ArenaCoding · as “claude-opus-4-5-20251101-high-32k”	Coding	1,495	100
44	LiveBenchReasoning · as “claude-opus-4-5-20251101-thinking-64k-high-effort”	Reasoning	72.6%	59
47	Epoch ECIReasoning · as “Claude Opus 4.5”	Reasoning	150.1	100
53	GPQA DiamondReasoning · as “claude-opus-4-5-20251101_32K”	Reasoning	86.0%	100
1	SWE-bench VerifiedCoding · stale since 2026-02-26 · as “Claude 4.5 Opus (high)”	Coding	76.8%	39
12	MMLU-ProReasoning · stale since 2026-03-11 · as “Claude-4.5-Opus(Thinking)”	Reasoning	87.3%	100

Category ranks and more models

Scores as published by each board on the capture date. Model names and logos belong to their owners; logos via logo.dev.