Design Arena: agents building Android apps leaderboard
- Grok 4.6 leads
- 1291
- models measured
- 39
- lead over #2
- +7 Elo
- median
- 1177
Leader on Design Arena: agents building Android apps · read October 22
On October 22, 2026, 39 of our catalog's models had a Design Arena: agents building Android apps result, and SpaceXAI Grok 4.6 led with 1291, ahead of Anthropic Claude Fable 5 at 1284.
Design Arena shows people two designs made by two models from the same prompt (a native Android app, built by the model as an agent) and asks which is better. The number is the Elo rating built from those votes, as Design Arena publishes it.
Role in the index
Design Arena: agents building Android apps is shown as published and does not feed any of our indexes. Read the methodology
| # | Model | Result | Battles | Evidence | Coding index | Output $/Mtok |
|---|---|---|---|---|---|---|
| 1 | Grok 4.6xAI | 1291 | Not published | 65.7 | $6.00 | |
| 2 | Claude Fable 5Anthropic | 1284 | 461 | 79.3 | $50.00 | |
| 3 | GLM 5V TurboZ.ai | 1267 | Not published | Not benchmarked | $4.00 | |
| 4 | Claude Opus 5Anthropic | 1257 | 397 | 83.0 | $25.00 | |
| 5 | Kimi K3Moonshot AI | 1253 | 460 | 62.3 | $15.00 | |
| 6 | Gemini 3.7 FlashGoogle | 1243 | 2,232 | 58.0 | $3.75 | |
| 7 | GPT-5.6 SolOpenAI | 1240 | 646 | 73.5 | $20.00 | |
| 8 | Grok 4.5xAI | 1236 | Not published | 58.6 | $6.00 | |
| 9 | Claude Opus 4.8Anthropic | 1233 | 949 | 65.0 | $25.00 | |
| 10 | Claude Sonnet 5Anthropic | 1227 | 956 | 57.7 | $10.00 | |
| 11 | Muse Spark 1.2Meta | 1223 | 846 | 55.7 | $4.25 | |
| 12 | Gemini 3.6 FlashGoogle | 1204 | 2,306 | 62.5 | $3.75 | |
| 13 | Claude Opus 4.6Anthropic | 1197 | 783 | 57.2 | $25.00 | |
| 14 | Claude Sonnet 4.6Anthropic | 1197 | 1,102 | 57.3 | $15.00 | |
| 15 | Qwen3.8 Max (0902)Qwen | 1193 | 151 | 62.4 | $6.00 | |
| 16 | Gemini 3.5 FlashGoogle | 1185 | 2,412 | 39.8 | $9.00 | |
| 17 | GLM 5.1Z.ai | 1182 | 1,082 | 49.2 | $4.40 | |
| 18 | Muse Spark 1.1Meta | 1178 | Not published | 70.3 | $4.25 | |
| 19 | GLM 5.2Z.ai | 1178 | 1,538 | 61.1 | $4.40 | |
| 20 | Claude Opus 4.5Anthropic | 1177 | 661 | 44.0 | $25.00 | |
| 21 | GPT-5.2-CodexOpenAI | 1176 | Not published | 45.4 | $14.00 | |
| 22 | GPT-5.5OpenAI | 1164 | 1,611 | 71.3 | $30.00 | |
| 23 | Kimi K2.6Moonshot AI | 1161 | 1,666 | 52.1 | $4.00 | |
| 24 | MiniMax M3MiniMax | 1159 | 1,174 | 39.2 | $1.20 | |
| 25 | Qwen3.7 MaxQwen | 1159 | 765 | 54.1 | $4.425 | |
| 26 | GLM 5Z.ai | 1147 | 1,602 | 41.7 | $3.20 | |
| 27 | Kimi K2.7 CodeMoonshot AI | 1138 | 658 | 53.4 | $4.00 | |
| 28 | Kimi K2.5Moonshot AI | 1119 | 554 | 38.6 | $2.85 | |
| 29 | GLM 4.7Z.ai | 1113 | 421 | 55.1 | $2.20 | |
| 30 | GLM 4.6Z.ai | 1085 | 500 | 8.8 | $2.20 | |
| 31 | GPT-5.3-CodexOpenAI | 1084 | Not published | 60.0 | $14.00 | |
| 32 | Gemini 3.8 FlashGoogle | 1079 | 337 | 61.6 | $3.75 | |
| 33 | GPT-5.6 TerraOpenAI | 1061 | 1,434 | 63.9 | $12.00 | |
| 34 | GPT-5.2OpenAI | 1060 | 660 | 55.9 | $14.00 | |
| 35 | Gemini 3 Flash PreviewGoogle | 1051 | 720 | 43.7 | $3.00 | |
| 36 | Gemini 3.1 Pro PreviewGoogle | 1050 | 5,369 | 32.3 | $12.00 | |
| 37 | Grok 4.20xAI | 1045 | Not published | 33.3 | $2.50 | |
| 38 | GPT-5.4OpenAI | 1032 | 502 | 64.4 | $15.00 | |
| 39 | Grok 4.3xAI | 962 | 1,815 | 37.4 | $2.50 |
Only models in our catalog are listed, and each appears once with its best published run. 146 rows were read from the source on October 22, 2026, and rows that match no catalog model are left out. Every other column is a figure the board publishes for the same run.