Anthropic

Claude Opus 4.6

Multimodal
#1Terminal Bench 2.0
#1ARC-AGI-2
#1SWE-rebench
+2 more

by Anthropic

+
+
+
+
About

Claude Opus 4.6, released by Anthropic in February 2026, builds on Opus 4.5 with improved coding precision, more careful planning, and greater ability to sustain agentic tasks over longer periods. It is designed for the most demanding autonomous workflows — scenarios requiring a model to operate independently through complex, multi-tool sequences without losing track of the overall objective.

+
+
+
+
Pricing Range
Input (per 1M)$5.00 -$5.00
Output (per 1M)$25.00 -$25.00
Providers3
+
+
+
+
Timeline
ReleasedFeb 5, 2026
Knowledge CutoffMay 1, 2025
+
+
+
+
Specifications
Capabilities
Multimodal
+
+
+
+
License & Family
License
Proprietary
Performance Overview
Performance metrics and category breakdown

Overall Performance

6 benchmarks
Average Score
69.7%
Best Score
91.3%
High Performers (80%+)
2

Performance Metrics

Max Context Window
328.0K

Top Categories

Science
91.3%
Reasoning
68.8%
Coding
66.0%
Finance
60.1%
+
+
+
+
All Benchmark Results for Claude Opus 4.6
Complete list of benchmark scores with detailed information
GPQA Diamond
Science
91.30
91.3%
Unverified
SWE Bench Verified
Coding
80.80
80.8%
Unverified
ARC-AGI-2
Reasoning
68.80
68.8%
Unverified
Terminal Bench 2.0
Coding
65.40
65.4%
Unverified
Finance Agent
Finance
60.10
60.1%
Unverified
SWE-rebench
Coding
51.70
51.7%
Unverified