ArtifactsBench — Published code/UI-artifact generation quality benchmark (2025)
| # | Model | Vendor | Best score | Runs | Last seen |
|---|---|---|---|---|---|
| 1 | GPT-5 | OpenAI | 73 | 1 | 2026-06-06 |
| 2 | MiniMax M2 | MiniMax | 66.8 | 1 | 2026-06-06 |
| 3 | Claude Sonnet 4.5 | Anthropic | 61.5 | 1 | 2026-06-06 |
| 4 | GLM-4.6 | Z.ai | 59.8 | 1 | 2026-06-06 |
| 5 | Gemini 2.5 Pro | 57.7 | 1 | 2026-06-06 | |
| 6 | Claude Sonnet 4 | Anthropic | 57.3 | 1 | 2026-06-06 |
| 7 | DeepSeek-V3.2 | DeepSeek | 55.8 | 1 | 2026-06-06 |
| 8 | Kimi K2 (Reasoning) | Moonshot | 54.2 | 1 | 2026-06-06 |
| 9 | MAI-Code-1-Flash | Microsoft | 36.4 | 1 | 2026-08-23 |
Best tracked score per model (default configuration; source-attributed and verification-tiered). Open a model for its full benchmark surface, provenance and pricing.