Claude Opus 4.1 Thinking

Released2025-08-05
Price per 1M$15.00 in · $75.00 out
Overall score50.3 · rank 259
Context window200,000 tokens
ProviderAnthropic
WeightsProprietary
ConfigurationThinking

Scores

DimensionScoreRankBenchmarks
Overall: Balanced 50.3 259 11
Overall: Coding 47.8 290 4
Overall: Reasoning 50.9 253 6
Overall: Agentic work 49.7 272 4
Overall: Frontend 46.9 260 3
Coding 46.2 290 1
Agents 51.1 194 1
Reasoning 49.9 264 2
Math 49.0 194 3
Knowledge 56.5 62 1
General 55.5 108 1
Multimodal 49.0 117 2
Value vs price−9.5342

Evidence

Coding 46.2 ±5.3

BenchmarkResultSourceDate
LiveCodeBench (Vals run) 66.5% Vals AI 2026-09-01

Agents 51.1 ±3.3

BenchmarkResultSourceDate
METR task suite 66.8% Epoch AI 2026-11-11

Reasoning 49.9 ±4.3

BenchmarkResultSourceDate
GPQA Diamond 77.3% Epoch AI 2025-08-05
GPQA Diamond (Vals run) 76.3% Vals AI 2026-09-01
Mystery Game Puzzles 21.0% Epoch AI 2026-07-25

Math 49.0 ±2.0

BenchmarkResultSourceDate
OTIS Mock AIME 2024-2025 68.9% Epoch AI 2025-08-05
FrontierMath Tier 4 2.4% Epoch AI 2026-06-11
FrontierMath Tiers 1-3 12.6% Epoch AI 2026-06-11
FrontierMath Tiers 1-3 (2025 set) 7.2% Epoch AI 2025-08-05
FrontierMath Tier 4 (2025 set) 4.2% Epoch AI 2025-08-05

Knowledge 56.5 ±4.7

BenchmarkResultSourceDate
MMLU-Pro (Vals run) 87.9% Vals AI 2026-09-01

General 55.5 ±4.1

BenchmarkResultSourceDate
LMArena Text 1,450 LMArena 2026-11-09

Multimodal 49.0 ±2.4

BenchmarkResultSourceDate
MMMU-Pro (Vals run) 77.5% Vals AI 2026-09-01
VPCT 33.0% Epoch AI 2026-11-11

Names matched to this model

claude-opus-4-1-20250805-thinking

Compare Claude Opus 4.1 Thinking with other models