← All benchmarks
Claude Sonnet 5 (Adaptive Reasoning, Max Effort)
Anthropic · Released 2026-06-30 · Independent benchmark scores from Artificial Analysis
Intelligence Index
38.4
Output speed
83.7
tokens/sec
Blended price
$4.00
per 1M tokens
Time to first token
122.96s
Model metadata
| Model ID | 23c86e4a-c769-43c0-a056-79e3cd15834f |
|---|---|
| Slug | claude-sonnet-5 |
| Release date | 2026-06-30 |
| Provider | Anthropic |
| Provider slug | anthropic |
| Output speed | 83.7 tok/s |
| Time to first token | 122.96s |
| Time to first answer | 122.96s |
| Blended price ($/1M) | $4.00 |
| Input price ($/1M) | $2.00 |
| Output price ($/1M) | $10.00 |
All evaluation scores (17)
| aime | — |
|---|---|
| aime 25 | — |
| artificial analysis coding index | 71.50 |
| artificial analysis intelligence index | 38.40 |
| artificial analysis math index | — |
| gpqa | 91.1% |
| hle | 41.3% |
| ifbench | — |
| lcr | 82.0% |
| livecodebench | — |
| math 500 | — |
| mmlu pro | — |
| scicode | 54.3% |
| tau banking | 37.3% |
| tau2 | — |
| terminalbench hard | — |
| terminalbench v2 1 | 80.5% |
Source: artificialanalysis.ai/models/claude-sonnet-5
Benchmark data from Artificial Analysis.