Best LLM for game development
- GPT-6 Astra leads
- 1474
- models ranked
- 127
- lead over #2
- +5 Elo
Design Arena: game dev · read October 21
As of October 21, 2026, OpenAI GPT-6 Astra leads 127 models on Design Arena: game dev with 1474, ahead of Anthropic Claude Sonnet 5.5 (1469) and Anthropic Claude Opus 5.5 (1457).
This list narrows the best LLM for web design ranking.
| Rank | Model | Design Arena: game dev | Evidence | Coding index | Output $/Mtok |
|---|---|---|---|---|---|
| 1 | GPT-6 Astra OpenAI | 1474 | 88.7 | $50.00 | |
| 2 | Claude Sonnet 5.5 Anthropic | 1469 | 88.9 | $10.00 | |
| 3 | Claude Opus 5.5 Anthropic | 1457 | 92.2 | $20.00 | |
| 4 | GPT-6.1 Sol OpenAI | 1389 | 88.7 | $10.00 | |
| 5 | Claude Fable 5.1 Anthropic | 1382 | 89.1 | $50.00 | |
| 6 | Muse Wonder Meta | 1380 | 64.5 | $20.00 | |
| 7 | Gemini 4 Argon Google | 1377 | 80.5 | $10.00 | |
| 8 | GPT-5.6 Sol OpenAI | 1368 | 76.7 | $20.00 | |
| 9 | DeepSeek V4.1 Flash 1017 DeepSeek | 1365 | 83.3 | $0.60 | |
| 10 | Kimi K3 Moonshot AI | 1361 | 61.8 | $11.936 | |
| 11 | Claude Opus 5 Anthropic | 1339 | 81.8 | $25.00 | |
| 12 | Claude Fable 5 Anthropic | 1336 | 82.1 | $50.00 | |
| 13 | DeepSeek V4.1 Flash DeepSeek | 1331 | 77.2 | $0.525 | |
| 14 | Muse Spark 1.3 Meta | 1329 | 79.7 | $4.25 | |
| 15 | Gemini 3.7 Flash Google | 1318 | 69.9 | $3.75 | |
| 16 | Claude Haiku 5.5 Anthropic | 1315 | 75.9 | $0.50 | |
| 17 | Grok 4.7 xAI | 1314 | 75.0 | $6.00 | |
| 18 | MiMo-V2.6-Pro Xiaomi | 1314 | 72.6 | $0.87 | |
| 19 | GPT-6 Sol OpenAI | 1313 | 82.8 | $10.00 | |
| 20 | Qwen3.8 Max (0902) Qwen | 1313 | 63.5 | $6.00 |
How this list is built
Generated from the dataset on every refresh. Nobody picks the order by hand.
Models by their Design Arena: game dev result, as the board publishes it, best first. Design Arena shows people two designs made by two models from the same prompt (a small game) and asks which is better. The number is the Elo rating built from those votes, as Design Arena publishes it.
- OpenRouter, fetched
- Design Arena, fetched