WebCoderBench - Performance: leaderboard
The Performance track of WebCoderBench — runtime performance of LLM-generated web front-end code; score (0-100).
Metric: Score (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 14 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3 Coder Plus | 98.54 | #338 |
| 2 | DeepSeek V3.1 Terminus | 98.19 | #220 |
| 3 | GPT-5 (High) | 98.08 | #98 (GPT-5) |
| 4 | DeepSeek V3.1 (Thinking) | 97.97 | #264 (DeepSeek V3.1) |
| 5 | DeepSeek V3.1 | 97.6 | #264 |
| 6 | Qwen 3 235B A22B Instruct | 97.45 | #333 |
| 7 | GLM-4.5 | 97.21 | #275 |
| 8 | MiniMax-M2 | 97.15 | #317 |
| 9 | DeepSeek R1 0528 | 97.02 | #222 |
| 10 | GPT-5 Codex (High) | 94.44 | #127 (GPT-5 Codex) |
| 11 | Gemini 2.5 Pro | 94.32 | #152 |
| 12 | GPT-4o (2024-11-20) | 94.07 | #381 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=webcoderbench-performance · How It Works · Data refreshed daily, snapshot 2026-10-20.