ArtifactsBenchmark: leaderboard
Evaluates LLM code generation for interactive artifacts across 1,825 tasks including UI components, games, SVGs, web apps, and simulations at 3 difficulty levels.
Metric: Average Score. Source: github.com. Source: benchmarklist.com. Saturation forecast: Around November 2027. 23 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Ling-3.0-flash | 77 | #206 |
| 2 | GPT-5 | 72.55 | #98 |
| 3 | Claude Opus 4.1 | 59.76 | #140 |
| 4 | Gemini 2.5 Pro | 57.74 | #152 |
| 5 | GPT-OSS-120B | 57.69 | #341 |
| 6 | Claude Sonnet 4 | 57.28 | #199 |
| 7 | Gemini 2.5 Pro (Preview 06-05) | 57.01 | #106 |
| 8 | Gemini 2.5 Pro (Preview 05-06) | 56.79 | #147 |
| 9 | Qwen 3 235B A22B (Thinking) | 55.01 | #313 (Qwen 3 235B A22B) |
| 10 | Claude 3.7 Sonnet | 52.19 | #249 |
| 11 | DeepSeek R1 0528 | 51.62 | #222 |
| 12 | GPT-4.1 | 48.23 | #248 |
| 13 | DeepSeek V3 (0324) | 45.56 | #340 |
| 14 | O3 Mini | 44.98 | #272 |
| 15 | DeepSeek R1 | 44.64 | #251 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=artifactsbenchmark · How It Works · Data refreshed daily, snapshot 2026-10-20.