PrepBench — leaderboard
Metric: End-to-End Prep-Code Acc. (self-reported). Source: benchmarklist.com. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.1 Codex | 54.9 |
| 2 | Gemini 3 Flash (Preview) | 53.3 |
| 3 | Claude Sonnet 4.5 | 52 |
| 4 | Kimi K2 (Thinking) | 49.7 |
| 5 | DeepSeek V3.2 | 44.8 |
| 6 | GLM-4.7 | 41.5 |
| 7 | Qwen 3 235B A22B | 38.2 |
| 8 | Grok Code Fast 1 | 30.1 |
| 9 | GPT-4o | 16.7 |
Interactive version: theaggregate.ai/benchmark?slug=prepbench · How the rankings work · Data refreshed daily, snapshot 2026-07-22.