Best AI model for UI components
In UI components, work within 60 Elo of the best model is available for 19× less: MiMo-V2.6-Pro at $0.472 per thousand tasks against $8.86 for Claude Opus 5.5.
106 models rated head to head in this category, not scored by a general benchmark and assumed to transfer. Single interface pieces built to spec, with states, accessibility and visual polish compared side by side.
Highest ratedClaude Opus 5.51390 Elo · $8.86 / 1k tasks
Best valueMiMo-V2.6-Pro1340 Elo · $0.472 / 1k tasks
Nothing cheaper is as good
5 models are undominated here: for each, no cheaper model scores within 18 Elo. Everything else in the table below is beaten on both counts at once.
- 1GPT-5 Nano1070 Elo$0.148 / 1k400K context
- 2Solar Pro 41154 Elo$0.164 / 1k524K context
- 3MiMo-V2.51260 Elo$0.183 / 1k1.05M context
- 4GLM 5.3 Flash1323 Elo$0.230 / 1k1.05M contextfp4 on some hosts
- 5Claude Sonnet 5.51378 Elo$4.03 / 1k1.00M context
Every model rated on UI components
Elo measured in this category specifically. A model strong overall can sit well down this table, which is the whole reason for ranking by task.
| # | Model | Elo | $/1k | Context | Hosts |
|---|---|---|---|---|---|
| 1 | Claude Opus 5.5 | 1390 | $8.86 | 1.00M | 11 |
| 2 | Claude Sonnet 5.5 | 1378 | $4.03 | 1.00M | 8 |
| 3 | Muse Spark 1.3 | 1356 | $1.98 | 1.05M | 1 |
| 4 | Kimi K3 | 1353 | $6.24 | 1.05M | 25 |
| 5 | Claude Opus 5 | 1349 | $11.25 | 1.00M | 11 |
| 6 | Claude Fable 5.1 | 1342 | $19.98 | 1.00M | 4 |
| 7 | MiMo-V2.6-Pro | 1340 | $0.472 | 1.05M | 4 |
| 8 | Gemini 3.8 Flash | 1323 | $1.54 | 1.05M | 6 |
| 9 | GLM 5.3 Flash | 1323 | $0.230 | 1.05M | 32 |
| 10 | Claude Fable 5 | 1322 | $20.52 | 1.00M | 6 |
| 11 | GLM 5.3 | 1321 | $1.93 | 1.05M | 41 |
| 12 | Muse Spark 1.2 | 1308 | $1.98 | 1.05M | 1 |
| 13 | Gemini 3.7 Flash | 1305 | $1.54 | 1.05M | 6 |
| 14 | Gemini 3.6 Flash | 1299 | $1.54 | 1.05M | 7 |
| 15 | Grok 4.6 | 1298 | $3.40 | 500K | 6 |
| 16 | GLM 5.2 | 1297 | $2.00 | 1.05M | 34 |
| 17 | Muse Spark 1.1 | 1294 | $1.98 | 1.05M | 1 |
| 18 | GLM 5.1 | 1292 | $2.11 | 203K | 13 |
| 19 | Claude Opus 4.6 | 1288 | $10.26 | 1.00M | 6 |
| 20 | Grok 4.5 | 1285 | $4.36 | 500K | 4 |
| 21 | Claude Sonnet 5 | 1283 | $4.30 | 1.00M | 10 |
| 22 | Gemini 3.1 Pro Preview | 1283 | $4.70 | 1.05M | 6 |
| 23 | Grok 4.7 | 1282 | $3.40 | 500K | 5 |
| 24 | Gemini 3.5 Flash | 1277 | $3.53 | 1.05M | 7 |
| 25 | Mistral Large 4 | 1276 | $1.00 | 1.05M | 3 |
| 26 | Claude Sonnet 4.6 | 1275 | $6.16 | 1.00M | 9 |
| 27 | Qwen3.7 Max | 1274 | $2.61 | 1.00M | 2 |
| 28 | Kimi K2.6 | 1269 | $1.52 | 262K | 17 |
| 29 | GLM 5 Turbo | 1267 | $1.95 | 203K | 1 |
| 30 | Qwen3.7 Plus | 1265 | $0.584 | 1.00M | 2 |
| 31 | Kimi K2.7 Code | 1263 | $1.75 | 262K | 12 |
| 32 | GPT-5.5 | 1262 | $12.90 | 1.05M | 7 |
| 33 | Claude Opus 4.8 | 1260 | $11.25 | 1.00M | 11 |
| 34 | MiMo-V2.5 | 1260 | $0.183 | 1.05M | 5 |
| 35 | MiMo-V2.5-Pro | 1251 | $0.497 | 1.05M | 6 |
| 36 | MiniMax M3 | 1250 | $0.547 | 1.05M | 13 |
| 37 | GPT-5.4 | 1243 | $6.45 | 1.05M | 7 |
| 38 | Claude Opus 4.5 | 1242 | $10.26 | 200K | 6 |
| 39 | GPT-5.1-Codex | 1241 | $3.69 | 400K | 1 |
| 40 | Kimi K2.5 | 1240 | $1.22 | 262K | 5 |
| 41 | DeepSeek V4 Flash 0731 | 1239 | $0.233 | 1.05M | 27 |
| 42 | Qwen3.6 Plus | 1236 | $0.975 | 1.00M | 1 |
| 43 | GLM 5 | 1234 | $1.42 | 203K | 8 |
| 44 | MiniMax M2.1 | 1226 | $0.706 | 205K | 2 |
| 45 | DeepSeek V4 Pro 0423 | 1225 | $1.68 | 1.05M | 15 |
| 46 | GLM 5V Turbo | 1216 | $1.95 | 203K | 1 |
| 47 | MiniMax M2.7 | 1213 | $0.534 | 197K | 6 |
| 48 | GLM 4.7 Flash | 1211 | $0.193 | 128K | 3 |
| 49 | Grok 4.20 | 1204 | $2.17 | 2.00M | 4 |
| 50 | GLM 4.7 | 1202 | $0.980 | 203K | 6 |
| 51 | Inkling | 1201 | $1.81 | 524K | 2 |
| 52 | Grok 4.3 | 1196 | $2.17 | 1.00M | 4 |
| 53 | GPT-5.2 | 1194 | $5.17 | 400K | 4 |
| 54 | DeepSeek V3.1 Terminus | 1186 | $0.672 | 164K | 2 |
| 55 | Step 3.7 Flash | 1185 | $0.470 | 262K | 2 |
| 56 | GPT-5 | 1183 | $3.69 | 400K | 3 |
| 57 | Qwen3.5 Plus 2026-02-15 | 1181 | $0.780 | 1.00M | 1 |
| 58 | MiniMax M2.5 | 1176 | $0.523 | 197K | 9 |
| 59 | Claude Sonnet 4.5 | 1176 | $6.16 | 1.00M | 7 |
| 60 | GPT-5.1 | 1170 | $3.69 | 400K | 5 |
| 61 | DeepSeek V3.2 Exp | 1170 | $0.447 | 164K | 1 |
| 62 | DeepSeek V4 Flash 0423 | 1168 | $0.170 | 1.05M | 16 |
| 63 | Claude Opus 4.1 | 1168 | $30.78 | 200K | 1 |
| 64 | Hy3 | 1167 | $0.265 | 262K | 5 |
| 65 | Qwen3.5 397B A17B | 1167 | $1.49 | 262K | 10 |
| 66 | Solar Pro 4 | 1154 | $0.164 | 524K | 2 |
| 67 | DeepSeek V3.2 | 1152 | $0.504 | 164K | 12 |
| 68 | GLM 4.6 | 1151 | $0.954 | 203K | 4 |
| 69 | GLM 4.5 | 1147 | $1.03 | 131K | 1 |
| 70 | Nemotron 3 Ultra | 1144 | $1.09 | 203K | 4 |
| 71 | Gemini 2.5 Pro | 1143 | $3.69 | 1.05M | 7 |
| 72 | GPT-5.3-Codex | 1141 | $5.17 | 400K | 3 |
| 73 | MiniMax M2 | 1140 | $0.666 | 205K | 2 |
| 74 | Claude Sonnet 4 | 1132 | $6.16 | 200K | 2 |
| 75 | GLM 4.5 Air | 1126 | $0.343 | 131K | 3 |
| 76 | Qwen3 Coder 480B A35B | 1116 | $0.666 | 262K | 3 |
| 77 | GPT-5 Mini | 1109 | $0.738 | 400K | 4 |
| 78 | R1 0528 | 1106 | $1.18 | 164K | 4 |
| 79 | Mistral Large 3 2512 | 1105 | $0.760 | 262K | 2 |
| 80 | Claude Haiku 4.5 | 1103 | $2.15 | 200K | 8 |
| 81 | Mistral Medium 3.1 | 1103 | $0.821 | 131K | 3 |
| 82 | DeepSeek V3 | 1097 | $0.634 | 128K | 2 |
| 83 | Gemini 2.5 Flash | 1096 | $0.916 | 1.05M | 7 |
| 84 | DeepSeek V3.1 | 1089 | $0.808 | 164K | 5 |
| 85 | GPT-5.1-Codex-Mini | 1084 | $0.742 | 400K | 1 |
| 86 | Gemini 3.1 Flash Lite Preview | 1075 | $0.588 | 1.05M | 3 |
| 87 | GPT-5 Nano | 1070 | $0.148 | 400K | 4 |
| 88 | Qwen3 Coder 30B A3B Instruct | 1049 | $0.264 | 262K | 2 |
| 89 | Trinity Large Thinking | 1048 | $0.403 | 262K | 1 |
| 90 | Kimi K2 0711 | 1034 | $1.37 | 131K | 1 |
| 91 | Mistral Medium 3 | 1028 | $0.821 | 131K | 3 |
| 92 | Codestral 2508 | 1020 | $0.456 | 256K | 2 |
| 93 | o3 | 1020 | $3.72 | 200K | 1 |
| 94 | GPT-4.1 | 1003 | $3.72 | 1.05M | 3 |
| 95 | Mercury 2 | 997 | $0.363 | 128K | 1 |
| 96 | o4 Mini | 987 | $2.05 | 200K | 1 |
| 97 | Qwen3 235B A22B Instruct 2507 | 968 | $0.405 | 262K | 7 |
| 98 | GPT-4.1 Mini | 964 | $0.744 | 1.05M | 3 |
| 99 | Qwen3 30B A3B | 949 | $0.294 | 41K | 1 |
| 100 | Qwen3 235B A22B Thinking 2507 | 948 | $1.59 | 128K | 1 |
| 101 | gpt-oss-120b | 930 | $0.285 | 131K | 24 |
| 102 | GPT-4.1 Nano | 919 | $0.186 | 1.05M | 3 |
| 103 | Mistral Small 3.2 24B | 910 | $0.193 | 128K | 4 |
| 104 | Llama 4 Maverick | 903 | $0.462 | 524K | 4 |
| 105 | GPT-4o | 889 | $5.10 | 128K | 2 |
| 106 | Llama 4 Scout | 771 | $0.393 | 328K | 3 |
Other tasks
Websitessave 4×Code generationsave 4×Game developmentsave 2×Data visualisationsave 4×3D graphicssave 2×SVG and vectorsave 39×ASCII artsave 6×
Elo comes from head-to-head build-offs via OpenRouter, compiled 12 October 2026 and refreshed daily. Cost is the median across every host serving each model, priced on 1,200 input and 300 output tokens with 60% cached. See the findings for method, or download the data.