Best AI model for 3D graphics
In 3D graphics, work within 60 Elo of the best model is available for 2× less: GPT-6.1 Sol at $4.43 per thousand tasks against $8.86 for Claude Opus 5.5.
103 models rated head to head in this category, not scored by a general benchmark and assumed to transfer. Three-dimensional scenes built in code: geometry, lighting and camera work, judged on whether the scene matches the brief.
Highest ratedClaude Opus 5.51482 Elo · $8.86 / 1k tasks
Best valueGPT-6.1 Sol1476 Elo · $4.43 / 1k tasks
Nothing cheaper is as good
7 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.
- 1Ministral 3 3B 2512966 Elo$0.085 / 1k131K context
- 2Ministral 3 8B 25121035 Elo$0.128 / 1k262K context
- 3Solar Pro 41182 Elo$0.164 / 1k524K context
- 4MiMo-V2.51220 Elo$0.183 / 1k1.05M context
- 5Claude Haiku 5.51351 Elo$0.215 / 1k1.00M context
- 6Muse Spark 1.31401 Elo$1.98 / 1k1.05M context
- 7GPT-6.1 Sol1476 Elo$4.43 / 1k1.05M context
Every model rated on 3D graphics
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 | 1482 | $8.86 | 1.00M | 11 |
| 2 | GPT-6.1 Sol | 1476 | $4.43 | 1.05M | 8 |
| 3 | Muse Spark 1.3 | 1401 | $1.98 | 1.05M | 1 |
| 4 | Claude Fable 5.1 | 1396 | $19.98 | 1.00M | 4 |
| 5 | Kimi K3 | 1394 | $6.24 | 1.05M | 25 |
| 6 | Claude Haiku 5.5 | 1351 | $0.215 | 1.00M | 10 |
| 7 | GLM 5.3 | 1342 | $1.93 | 1.05M | 41 |
| 8 | Claude Opus 5 | 1337 | $11.25 | 1.00M | 11 |
| 9 | GLM 5.1 | 1336 | $2.11 | 203K | 13 |
| 10 | MiMo-V2.6-Pro | 1335 | $0.472 | 1.05M | 4 |
| 11 | Claude Fable 5 | 1322 | $20.52 | 1.00M | 6 |
| 12 | GLM 5.3 Flash | 1321 | $0.230 | 1.05M | 32 |
| 13 | Gemini 3.7 Flash | 1311 | $1.54 | 1.05M | 6 |
| 14 | Muse Spark 1.2 | 1306 | $1.98 | 1.05M | 1 |
| 15 | Gemini 3.8 Flash | 1299 | $1.54 | 1.05M | 6 |
| 16 | GLM 5.2 | 1292 | $2.00 | 1.05M | 34 |
| 17 | Grok 4.6 | 1280 | $3.40 | 500K | 6 |
| 18 | Gemini 3.6 Flash | 1277 | $1.54 | 1.05M | 7 |
| 19 | Claude Opus 4.6 | 1277 | $10.26 | 1.00M | 6 |
| 20 | Kimi K2.6 | 1271 | $1.52 | 262K | 17 |
| 21 | Qwen3.7 Max | 1270 | $2.61 | 1.00M | 2 |
| 22 | Grok 4.7 | 1269 | $3.40 | 500K | 5 |
| 23 | Claude Sonnet 5 | 1265 | $4.30 | 1.00M | 10 |
| 24 | Muse Spark 1.1 | 1264 | $1.98 | 1.05M | 1 |
| 25 | Grok 4.5 | 1262 | $4.36 | 500K | 4 |
| 26 | DeepSeek V4 Pro 0423 | 1257 | $1.68 | 1.05M | 15 |
| 27 | Mistral Large 4 | 1253 | $1.00 | 1.05M | 3 |
| 28 | GLM 5 Turbo | 1253 | $1.95 | 203K | 1 |
| 29 | Gemini 3.5 Flash | 1248 | $3.53 | 1.05M | 7 |
| 30 | MiMo-V2.5-Pro | 1248 | $0.497 | 1.05M | 6 |
| 31 | Kimi K2.7 Code | 1245 | $1.75 | 262K | 12 |
| 32 | Qwen3.7 Plus | 1241 | $0.584 | 1.00M | 2 |
| 33 | Gemini 3.1 Pro Preview | 1238 | $4.70 | 1.05M | 6 |
| 34 | Claude Sonnet 4.6 | 1238 | $6.16 | 1.00M | 9 |
| 35 | GLM 5 | 1232 | $1.42 | 203K | 8 |
| 36 | Claude Opus 4.8 | 1226 | $11.25 | 1.00M | 11 |
| 37 | MiMo-V2.5 | 1220 | $0.183 | 1.05M | 5 |
| 38 | Claude Opus 4.5 | 1217 | $10.26 | 200K | 6 |
| 39 | DeepSeek V4 Flash 0731 | 1212 | $0.233 | 1.05M | 27 |
| 40 | MiniMax M3 | 1212 | $0.547 | 1.05M | 13 |
| 41 | GLM 5V Turbo | 1212 | $1.95 | 203K | 1 |
| 42 | Kimi K2.5 | 1208 | $1.22 | 262K | 5 |
| 43 | Qwen3.6 Plus | 1205 | $0.975 | 1.00M | 1 |
| 44 | GPT-5.5 | 1201 | $12.90 | 1.05M | 7 |
| 45 | MiniMax M2.7 | 1197 | $0.534 | 197K | 6 |
| 46 | Grok 4.20 | 1196 | $2.17 | 2.00M | 4 |
| 47 | GLM 4.7 | 1194 | $0.980 | 203K | 6 |
| 48 | DeepSeek V4 Flash 0423 | 1192 | $0.170 | 1.05M | 16 |
| 49 | Gemini 3 Flash Preview | 1187 | $1.18 | 1.05M | 6 |
| 50 | Solar Pro 4 | 1182 | $0.164 | 524K | 2 |
| 51 | Hy3 | 1179 | $0.265 | 262K | 5 |
| 52 | GLM 4.5 | 1175 | $1.03 | 131K | 1 |
| 53 | MiniMax M2.5 | 1173 | $0.523 | 197K | 9 |
| 54 | MiniMax M2.1 | 1168 | $0.706 | 205K | 2 |
| 55 | Inkling | 1166 | $1.81 | 524K | 2 |
| 56 | Qwen3.5 397B A17B | 1164 | $1.49 | 262K | 10 |
| 57 | Claude Sonnet 4.5 | 1161 | $6.16 | 1.00M | 7 |
| 58 | Claude Opus 4.1 | 1156 | $30.78 | 200K | 1 |
| 59 | DeepSeek V3.2 Exp | 1153 | $0.447 | 164K | 1 |
| 60 | DeepSeek V3.1 Terminus | 1144 | $0.672 | 164K | 2 |
| 61 | Claude Sonnet 4 | 1143 | $6.16 | 200K | 2 |
| 62 | DeepSeek V3.2 | 1136 | $0.504 | 164K | 12 |
| 63 | Nemotron 3 Ultra | 1135 | $1.09 | 203K | 4 |
| 64 | Step 3.7 Flash | 1132 | $0.470 | 262K | 2 |
| 65 | GLM 4.6 | 1132 | $0.954 | 203K | 4 |
| 66 | Grok 4.3 | 1128 | $2.17 | 1.00M | 4 |
| 67 | GLM 4.5 Air | 1126 | $0.343 | 131K | 3 |
| 68 | GLM 4.7 Flash | 1124 | $0.193 | 128K | 3 |
| 69 | Qwen3.5 Plus 2026-02-15 | 1122 | $0.780 | 1.00M | 1 |
| 70 | R1 0528 | 1114 | $1.18 | 164K | 4 |
| 71 | GPT-5.4 | 1105 | $6.45 | 1.05M | 7 |
| 72 | Mistral Large 3 2512 | 1103 | $0.760 | 262K | 2 |
| 73 | MiniMax M2 | 1101 | $0.666 | 205K | 2 |
| 74 | DeepSeek V3 | 1090 | $0.634 | 128K | 2 |
| 75 | Mistral Medium 3 | 1087 | $0.821 | 131K | 3 |
| 76 | Trinity Large Thinking | 1085 | $0.403 | 262K | 1 |
| 77 | Mistral Medium 3.1 | 1085 | $0.821 | 131K | 3 |
| 78 | Gemini 2.5 Pro | 1084 | $3.69 | 1.05M | 7 |
| 79 | GPT-5.2 | 1082 | $5.17 | 400K | 4 |
| 80 | DeepSeek V3.1 | 1080 | $0.808 | 164K | 5 |
| 81 | Claude Haiku 4.5 | 1077 | $2.15 | 200K | 8 |
| 82 | Gemini 2.5 Flash | 1074 | $0.916 | 1.05M | 7 |
| 83 | GPT-5.1 | 1064 | $3.69 | 400K | 5 |
| 84 | GPT-5 | 1059 | $3.69 | 400K | 3 |
| 85 | Gemini 3.1 Flash Lite Preview | 1051 | $0.588 | 1.05M | 3 |
| 86 | GPT-5 Mini | 1040 | $0.738 | 400K | 4 |
| 87 | Ministral 3 8B 2512 | 1035 | $0.128 | 262K | 3 |
| 88 | Codestral 2508 | 1023 | $0.456 | 256K | 2 |
| 89 | GPT-5.3-Codex | 1011 | $5.17 | 400K | 3 |
| 90 | Qwen3 235B A22B Thinking 2507 | 1001 | $1.59 | 128K | 1 |
| 91 | Qwen3 235B A22B Instruct 2507 | 999 | $0.405 | 262K | 7 |
| 92 | Ministral 3 14B 2512 | 993 | $0.170 | 262K | 3 |
| 93 | GPT-5.1-Codex-Mini | 992 | $0.742 | 400K | 1 |
| 94 | Mercury 2 | 984 | $0.363 | 128K | 1 |
| 95 | GPT-5 Nano | 969 | $0.148 | 400K | 4 |
| 96 | Ministral 3 3B 2512 | 966 | $0.085 | 131K | 3 |
| 97 | GPT-4.1 Nano | 928 | $0.186 | 1.05M | 3 |
| 98 | gpt-oss-120b | 906 | $0.285 | 131K | 24 |
| 99 | Llama 4 Maverick | 904 | $0.462 | 524K | 4 |
| 100 | GPT-4o | 872 | $5.10 | 128K | 2 |
| 101 | o4 Mini | 857 | $2.05 | 200K | 1 |
| 102 | GPT-4.1 | 853 | $3.72 | 1.05M | 3 |
| 103 | GPT-4.1 Mini | 840 | $0.744 | 1.05M | 3 |
Other tasks
Websitessave 4×Code generationsave 4×Game developmentsave 2×Data visualisationsave 4×UI componentssave 19×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.