Claude Opus 4.5
Multimodal
#1SWE Bench Verified
#1MCP-Atlas
#2GDPVal
+4 more
by Anthropic
+
+
+
+
About
Claude Opus 4.5, released by Anthropic in November 2025, is a large language model from the Claude 4.5 family built for demanding reasoning tasks, advanced code generation, and complex agentic workflows. It features a 200K token context window, 64K maximum output tokens, native image understanding, and extended thinking with configurable effort levels. Opus 4.5 targets deep analytical work, multi-step tool orchestration, and applications requiring sustained reasoning across long, complex tasks.
+
+
+
+
Pricing Range
Input (per 1M)$5.00 -$5.00
Output (per 1M)$25.00 -$25.00
Providers3
+
+
+
+
Timeline
ReleasedNov 1, 2025
Knowledge CutoffMay 1, 2025
+
+
+
+
Specifications
Capabilities
Multimodal
+
+
+
+
License & Family
License
Proprietary
Performance Overview
Performance metrics and category breakdown
Overall Performance
19 benchmarks
Average Score
61.9%
Best Score
98.2%
High Performers (80%+)
5Performance Metrics
Max Context Window
264.0KTop Categories
Knowledge
90.8%
Science
87.0%
Agents
72.9%
Tool Use
62.3%
Coding
61.5%
+
+
+
+
All Benchmark Results for Claude Opus 4.5
Complete list of benchmark scores with detailed information
| TAU2-Bench Telecom | Agents | 98.20 | 98.2% | Self-reported | |
| MMMLU | Knowledge | 90.80 | 90.8% | Self-reported | |
| TAU2-Bench Retail | Agents | 88.90 | 88.9% | Self-reported | |
| GPQA Diamond | Science | 87.00 | 87.0% | Unverified | |
| SWE Bench Verified | Coding | 80.90 | 80.9% | Unverified | |
| MMMU-Pro with Tools | Multimodal | 73.90 | 73.9% | Self-reported | |
| GDPVal AA ELO | Agents | 1416.00 | 70.8% | Self-reported | |
| MMMU-Pro | Multimodal | 70.60 | 70.6% | Self-reported | |
| BrowseComp | Agents | 67.80 | 67.8% | Self-reported | |
| OSWorld | Agents | 66.30 | 66.3% | Self-reported |
Showing 1 to 10 of 19 benchmarks