Best coding models for design
This page uses the WebDev Arena benchmark as a frontend/design signal, then adds current token price, context, and host availability. It is not a general visual-design quality ranking: the benchmark source is an external dataset and remains non-commercial until its upstream licence is confirmed.
| # | Model | WebDev score | Input $/MTok | Output $/MTok | Hosts | Context |
|---|---|---|---|---|---|---|
| 1 | Claude Opus 5 | 1687.0 | $5.00 | $25.00 | 1 | 1.0M |
| 2 | Kimi K3 | 1674.3 | $3.00 | $15.00 | 2 | — |
| 3 | Claude Fable 5 | 1626.3 | $10.00 | $50.00 | 1 | 1.0M |
| 4 | GLM 5.3 | 1614.5 | $2.10 | $6.60 | 1 | 1.0M |
| 5 | GLM 5.3 Flash | 1607.1 | $0.23 | $0.75 | 1 | 1.0M |
| 6 | Qwen3.8 27b | 1593.3 | $0.40 | $2.37 | 3 | — |
| 7 | GLM-5.2 | 1593.3 | $1.20 | $4.20 | 4 | — |
| 8 | Claude Opus 4.7 | 1566.8 | $5.00 | $25.00 | 1 | 1.0M |
| 9 | Claude Opus 4.6 | 1556.3 | $5.00 | $25.00 | 1 | 1.0M |
| 10 | Claude Opus 4.8 | 1552.2 | $5.00 | $25.00 | 1 | 1.0M |
| 11 | Gemini 3.6 Flash | 1527.8 | $0.75 | $3.75 | 1 | 1.0M |
| 12 | Claude Sonnet 4.6 | 1521.0 | $3.00 | $15.00 | 1 | 1.0M |
| 13 | Kimi-K2.6 | 1508.8 | $0.73 | $3.50 | 3 | 262k |
| 14 | GLM 5.1 | 1508.3 | $1.40 | $4.40 | 1 | — |
| 15 | Gemini 3.5 Flash | 1506.4 | $1.50 | $9.00 | 1 | 1.0M |
| 16 | MiniMax M3 | 1486.7 | $0.30 | $1.20 | 3 | — |
| 17 | Kimi K2.7 Code | 1472.7 | $0.75 | $3.50 | 4 | 262k |
| 18 | DeepSeek V4 Pro | 1463.6 | $1.32 | $3.50 | 5 | — |
| 19 | GPT-5.5 | 1457.8 | $5.00 | $30.00 | 1 | 1.1M |
| 20 | Gemini 3.5 Flash Lite | 1446.6 | $0.30 | $2.50 | 1 | 1.0M |
| 21 | Kimi K2.5 | 1436.4 | $0.50 | $2.80 | 7 | — |
| 22 | GLM 5 | 1435.9 | $1.20 | $3.84 | 6 | — |
| 23 | GLM 4.7 | 1434.7 | $0.72 | $2.64 | 6 | — |
| 24 | MiniMax M2.7 | 1397.5 | $0.30 | $1.20 | 1 | 197k |
| 25 | Minimax M2.1 | 1387.5 | $0.36 | $1.44 | 6 | — |