Step 5 Preview (High): benchmark results
Provider: StepFun. Released 2026-09-20. Access: API.
Aggregate Rating 1778 ± 23, rank #195 of 2245 rated models, from 72 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| FrontierCS 2 - OCR Distillation Trajectory Sampling | 85.45 | Task score (0–100) | 94.7 |
| SEAL - SWE Atlas - Codebase QnA | 63.6 | Score | 90.6 |
| FrontierCS 2 - EVM State Checkpoints | 66.92 | Task score (0–100) | 89.5 |
| SEAL - MCP Atlas | 85.6 | Score | 88.7 |
| AA CritPt | 20.9 | Accuracy (%) | 88.6 |
| FrontierCS 2 - In-place Integer Redistribution | 51.07 | Task score (0–100) | 78.9 |
| WebDev Arena (Brand & Marketing) | 1590 | Arena Score | 78.4 |
| WebDev Arena | 1570.15 | Arena Score | 78 |
| WebDev Arena (Gaming) | 1608 | Arena Score | 77 |
| WebDev Arena (Frontend) | 1576 | Arena Score | 75.7 |
| Chatbot Arena (Code) | 1564 | Elo | 75.2 |
| WebDev Arena (Simulations) | 1571 | Arena Score | 74.1 |
Interactive version: theaggregate.ai/model?slug=step-5-preview-high · How It Works · Data refreshed daily, snapshot 2026-11-16.