WebCoderBench - Functionality Correctness: leaderboard
The Functionality Correctness track of WebCoderBench — whether LLM-generated web front-end code implements correctly working features; score (0-100).
Metric: Score (%). Source: huggingface.co. Saturation forecast: Around January 2027. 14 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5 (High) | 79.18 | #98 (GPT-5) |
| 2 | GPT-5 Codex (High) | 74.82 | #127 (GPT-5 Codex) |
| 3 | MiniMax-M2 | 72.52 | #317 |
| 4 | Gemini 2.5 Pro | 68.28 | #152 |
| 5 | GLM-4.5 | 57.73 | #275 |
| 6 | Qwen 3 235B A22B Instruct | 56.41 | #333 |
| 7 | DeepSeek V3.1 Terminus | 56.25 | #220 |
| 8 | Qwen 3 Coder Plus | 54.95 | #338 |
| 9 | DeepSeek R1 0528 | 54.67 | #222 |
| 10 | DeepSeek V3.1 (Thinking) | 54.03 | #264 (DeepSeek V3.1) |
| 11 | DeepSeek V3.1 | 53.68 | #264 |
| 12 | GPT-4o (2024-11-20) | 49.42 | #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-functionality-correctness · How It Works · Data refreshed daily, snapshot 2026-10-20.