Claude Sonnet 4
Multimodal
Zero-eval
by Anthropic
+
+
+
+
About
Claude Sonnet 4 is a multimodal language model developed by Anthropic. It achieves strong performance with an average score of 69.4% across 8 benchmarks. It excels particularly in MMMLU (86.5%), TAU-bench Retail (80.5%), GPQA (75.4%). It supports a 328K token context window for handling large documents. The model is available through 4 API providers. As a multimodal model, it can process and understand text, images, and other input formats seamlessly. Released in 2025, it represents Anthropic's latest advancement in AI technology.
+
+
+
+
Pricing Range
Input (per 1M)$3.00 -$3.00
Output (per 1M)$15.00 -$15.00
Providers4
+
+
+
+
Timeline
AnnouncedMay 22, 2025
ReleasedMay 22, 2025
+
+
+
+
Specifications
Capabilities
Multimodal
+
+
+
+
License & Family
License
Proprietary
Performance Overview
Performance metrics and category breakdown
Overall Performance
8 benchmarks
Average Score
69.4%
Best Score
86.5%
High Performers (80%+)
2Performance Metrics
Max Context Window
328.0KAvg Throughput
56.8 tok/sAvg Latency
0ms+
+
+
+
All Benchmark Results for Claude Sonnet 4
Complete list of benchmark scores with detailed information
| MMMLU | text | 0.86 | 86.5% | Self-reported | |
| TAU-bench Retail | text | 0.81 | 80.5% | Self-reported | |
| GPQA | text | 0.75 | 75.4% | Self-reported | |
| MMMU | multimodal | 0.74 | 74.4% | Self-reported | |
| SWE-Bench Verified | text | 0.73 | 72.7% | Self-reported | |
| AIME 2025 | text | 0.70 | 70.5% | Self-reported | |
| TAU-bench Airline | text | 0.60 | 60.0% | Self-reported | |
| Terminal-Bench | text | 0.35 | 35.5% | Self-reported |