Anthropic model product
5 Sonnet is a powerful AI model.
Updated Aug 12, 2026. Default version: Claude 3.5 Sonnet
Technical details for the model's default version.
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Benchmark scores for Claude 3.5 Sonnet.
| AI2D | 94.7% | 01 | 32 | 100.0% | C | |
| ChartQA | 90.8% | 01 | 24 | 100.0% | C | |
| OSWorld Extended | 22.0% | 01 | 1 | 100.0% | C | |
| OSWorld Screenshot-only | 14.9% | 01 | 1 | 100.0% | C | |
| BIG-Bench Hard | 93.1% | 02 | 21 | 95.0% | C | |
| DocVQA | 95.2% | 03 | 26 | 92.0% | C | |
| DROP | 87.1% | 03 | 30 | 93.1% | C | |
| HumanEval | 93.7% | 03 | 66 | 96.9% | C | |
| MGSM | 91.6% | 04 | 31 | 90.0% | C | |
| GSM8k | 96.4% | 07 | 48 | 87.2% | C | |
| MMLU | 90.4% | 08 | 100 | 92.9% | C | |
| TAU-bench Retail | 69.2% | 12 | 25 | 54.2% | C | |
| TAU-bench Airline | 46.0% | 16 | 23 | 31.8% | C | |
| MathVista | 67.7% | 20 | 39 | 50.0% | C | |
| MATH | 78.3% | 21 | 71 | 71.4% | C | |
| MMMU | 68.3% | 35 | 63 | 45.2% | C | |
| MMLU-Pro | 77.6% | 62 | 129 | 52.3% | C | |
| SWE-Bench Verified | 49.0% | 90 | 105 | 14.4% | C | |
| GPQA | 67.2% | 138 | 234 | 41.2% | C |
Preference and agent-evaluation results for the default version.
| vision style control | overall | 90 | 1161.1 | 10,389 | N/A | |
| vision | overall | 104 | 1125.5 | 10,389 | N/A | |
| text style control | overall | 169 | 1373.4 | 88,405 | N/A | |
| text | overall | 222 | 1297.8 | 88,405 | N/A |
Provider-specific output speed and catalog latency for Claude 3.5 Sonnet. Runtime does not affect the capability score.
| Bedrock | 101 tok/s | 0.5 s | 200K | 200K | |
| Anthropic | 100 tok/s | 0.5 s | 200K | 200K | |
| 42 tok/s | 0.4 s | 200K | 200K |
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
Official vendor API pricing appears first, followed by individual provider offers.
The default version has no current input or output token prices.
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Available versions of this model. The score column identifies the version used in the overall ranking.
| Claude 3.5 Sonnet | 32.1 | N/A | 200K | 8.2K | No | Proprietary | |
| Claude 3.5 Sonnet | N/A | N/A | 200K | 200K | No | Proprietary |
Key information about Claude Sonnet 3.5 and its available data.
5 Sonnet is a powerful AI model. It excels in graduate-level reasoning, undergraduate-level knowledge, and coding proficiency, with improved understanding of nuance, humor, and complex instructions.
Data as of 2026-08-11.
Open a comparison with the three ranked models immediately above and below this model.
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
Common questions about Claude Sonnet 3.5.
Claude Sonnet 3.5's default version was released on Oct 22, 2024.
No official standard PAYG price is currently available for Claude Sonnet 3.5.
Claude Sonnet 3.5 was created by Anthropic.
The default version has a 200K token context window.
No. The default version is not marked as having publicly available weights.
No provider offering is currently linked to the default version.
Nearby ranked alternatives include LongCat Flash Lite, Qwen3.5 4B, Nemotron Nano 9B.