Anthropic: Claude Sonnet 4.6 wins on more metrics (5 of 8), but the right pick depends on what you optimize for — see the breakdown below.
| Metric | Anthropic: Claude Sonnet 4.6 | StepFun: Step 3.7 Flash |
|---|---|---|
| Intelligence Index | 42.6 | 42.6 |
| Coding Index | 43.0✓ | 37.1 |
| GPQA Diamond | 80% | 81%✓ |
| Design Arena Elo | 1328✓ | 1232 |
| Speed (tokens/sec) | 46✓ | — |
| Latency | 1.1s✓ | — |
| Input price /M | $3.00 | $0.200✓ |
| Output price /M | $15.00 | $1.15✓ |
| Context window | 1M✓ | 256K |
| Capabilities | ReasoningToolsJSONVision | ReasoningToolsJSONVision |