Inference
EN
Anthropic

Claude Opus 4

Claude Opus 4 is a model from Anthropic. 14 benchmarks count toward its index, in 6 categories.

  • Input: text
RankedEnough results for an index.
IndexAll categories, out of 100.43.9±5.3Median 51.0
CoverageIndex weight that has results.90%
SpeedOutput tokens per second.—Median 67/s
First tokenSeconds to the first output token.—Median 1.3 s
Input / 1MUS dollars per 1M input tokens.—Median $0.60
Output / 1MUS dollars per 1M output tokens.—Median $2.55
ContextMaximum tokens in one request.—Median 262K
EloLMArena rating and rank.1366±4 #195Median 1422

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.1 counted49.6—
CodingCode writing and repair.2 counted47.3153 / 209
ReasoningLogic problems and puzzles.2 counted38.6172 / 227
MultimodalTasks with images and text.1 counted50.6—
KnowledgeFacts and expert knowledge.2 counted48.2147 / 268
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.None counted——
MathMath problems.6 counted39.4126 / 168

Results

14 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
MATH level 5Math85.0% ±1.022 / 88
MMLU ProKnowledge86.2% ±0.345 / 135
τ²-bench RetailAgentic81.4%8 / 23
MATH 500Math90.4% ±1.324 / 59
GPQA diamond · Epoch AIKnowledge72.7% ±3.0109 / 217
OTIS Mock AIME 2024-2025Math60.0% ±7.4112 / 200
τ²-bench AirlineAgentic59.6%14 / 22
FrontierMath-Tier-4-2025-07-01-PrivateMath2.1%35 / 55
MMMU ProMultimodal73.3% ±1.160 / 92
GPQA Diamond · Vals AIKnowledge71.7% ±2.390 / 135
FrontierMath-2025-02-28-PrivateMath4.3% ±1.248 / 69
LiveCodeBenchCoding62.6% ±1.1103 / 142
AIMEMath41.3% ±1.269 / 95
SWE-Bench verified · Epoch AICoding70.7% ±2.124 / 32
ARC-AGI-2 (semi-private)Reasoning1.3%103 / 131
ARC-AGI-1 (semi-private)Reasoning22.5%103 / 126
FrontierMath-Tier-4-2025-07-01-PublicMath0.0%3 / 36
MGSMMultilingual93.8% ±1.58 / 74
MedQAHealthcare92.9% ±0.227 / 95
LegalBenchLegal83.1% ±0.458 / 149
TaxEval v2Finance71.9% ±0.970 / 143
FrontierMath-2025-02-28-PublicMath0.0%33 / 64
MortgageTaxFinance58.6% ±1.069 / 97
ARC-AGI-2 (public eval)Reasoning1.3%91 / 122
ARC-AGI-1 (public eval)Reasoning35.5%92 / 116

Sources

BenchLMCC BY-NC 4.0Vals AINo licence statedEpoch AICC BYSierra ResearchMITARC Prize FoundationTerms restrict reuse

History

5 releases
Index44.9 → 43.9
23 Oct 202612 Nov 2026

Questions

How good is Claude Opus 4?
Its index is 43.9 out of 100. That is place 186 of 283 ranked models.
What is Claude Opus 4 best at?
Knowledge: 48.2 out of 100, place 147 of 268.

Badge

Claude Opus 4 on Inference360: place 186 of 283, index 43.9
HTML
<a href="https://inference360.ai/models/anthropic-claude-opus-4"><img src="https://inference360.ai/badges/anthropic-claude-opus-4.svg" alt="Claude Opus 4 on Inference360: place 186 of 283, index 43.9" height="20"></a>
Markdown
[![Claude Opus 4 on Inference360: place 186 of 283, index 43.9](https://inference360.ai/badges/anthropic-claude-opus-4.svg)](https://inference360.ai/models/anthropic-claude-opus-4)

Compare with

DeepSeek V3.244.2Claude Haiku 4.5 Thinking43.5

More from Anthropic

All models
Claude Fable 5.285.6Claude Opus 5.578.6Claude Sonnet 5.577.2Claude Fable 5.176.2Claude Fable 574.0