Ranking · código
¿Qué modelo programa mejor?
WebDev Arena enfrenta a dos modelos construyendo la misma aplicación web y la gente vota cuál funciona mejor. Lo cruzamos con el Coding Index, el Intelligence Index, el precio y la velocidad de Artificial Analysis. Leaderboard del 16 de octubre de 2026.
Elo en duelos construyendo aplicaciones web
| # | Modelo | WebDev Arena | Coding Index | $/M tokens |
|---|---|---|---|---|
| 1 | Claude Opus 5.5 (Max)Pareto anthropicProprietary | 1807 | · | $8 |
| 2 | GPT-6 Astra (Max) openaiProprietary | 1777 | 76,9 | $20 |
| 3 | Claude Sonnet 5.5 (xHigh)Pareto anthropicProprietary | 1768 | · | $4 |
| 4 | GPT-6.1 Sol (Max) openaiProprietary | 1755 | · | $4 |
| 5 | Claude Fable 5.1 (Max) anthropicProprietary | 1741 | 81,6 | $20 |
| 6 | Claude Sonnet 5.5 (High) anthropicProprietary | 1711 | · | $4 |
| 7 | Claude Opus 5 (Max) anthropicProprietary | 1689 | 78,0 | $10 |
| 8 | GPT-6 Sol (Max) openaiProprietary | 1683 | · | $4 |
| 9 | Gemini 4 Argon (High) googleProprietary | 1678 | · | $4 |
| 10 | Qwen3.8-Max (0902)Pareto alibabaProprietary | 1674 | 76,2 | $3 |
| 11 | Qwen3.8-Max alibabaProprietary | 1672 | 71,8 | $3 |
| 12 | Muse Spark 1.3 (Max)Pareto metaProprietary | 1658 | 75,8 | $2 |
| 13 | Claude Opus 5 (High) anthropicProprietary | 1657 | 76,5 | $10 |
| 14 | Kimi K3 (Max) moonshotKimi K3 license | 1655 | 76,2 | $6 |
| 15 | Muse Wonder (Max) metaProprietary | 1650 | · | $8 |
| 16 | Grok 4.7 (xHigh) xaiProprietary | 1639 | · | $3 |
| 17 | MiMo-V2.6-FlashPareto xiaomiMIT | 1634 | · | $0,18 |
| 18 | Hy4 preview tencentApache 2.0 | 1631 | · | · |
| 19 | Qwen3.8-Flash-Next alibabaqwen-community-1.0 | 1630 | 73,1 | $0,23 |
| 20 | MiMo-V2.6-Pro xiaomiMIT | 1625 | · | $0,54 |
| 21 | Claude Fable 5 (High) anthropicProprietary | 1625 | · | · |
| 22 | Muse Spark 1.3 (xHigh) metaProprietary | 1622 | 76,5 | $2 |
| 23 | GLM-5.3 (Max) zaiMIT | 1621 | 74,8 | $2,15 |
| 24 | DeepSeek-V4.1-Flash (Max) deepseekMIT | 1618 | · | $0,53 |
| 25 | Grok 4.6 (High) xaiProprietary | 1617 | 76,8 | $3 |
| 26 | GPT-5.6 Sol (xHigh, Codex Harness) openaiProprietary | 1616 | · | · |
| 27 | GLM-5.3-Flash zaiMIT | 1605 | 71,5 | $0,24 |
| 28 | GLM-5.2 (Max) zaiMIT | 1605 | 68,8 | $2,15 |
| 29 | Qwen3.8-27B alibabaApache 2.0 | 1593 | · | · |
| 30 | Gemini 3.7 Flash (High) googleProprietary | 1591 | 76,1 | $1,5 |
| 31 | Gemini 3.8 Flash (High) googleProprietary | 1585 | 76,3 | $1,5 |
| 32 | Claude Haiku 5.5 (High) anthropicProprietary | 1585 | · | $0,2 |
| 33 | DeepSeek-V4-Pro (High, 20260813) deepseekMIT | 1582 | · | · |
| 34 | DeepSeek-V4-Flash (High) deepseekMIT | 1581 | · | · |
| 35 | GPT-6 Luna (Max) openaiProprietary | 1578 | · | $0,2 |
| 36 | Step 5 Preview (High) stepfunProprietary | 1565 | · | · |
| 37 | Claude Opus 4.7 (High) anthropicProprietary | 1558 | · | · |
| 38 | Claude Opus 4.7 anthropicProprietary | 1556 | · | · |
| 39 | Claude Opus 4.8 (High) anthropicProprietary | 1556 | · | · |
| 40 | Grok 4.5 xaiProprietary | 1553 | · | · |
| 41 | Claude Opus 4.6 (High) anthropicProprietary | 1547 | · | · |
| 42 | Muse Spark 1.1 metaProprietary | 1542 | 71,3 | $2 |
| 43 | Mistral Large 4 mistralProprietary | 1542 | · | · |
| 44 | Claude Sonnet 5 (High) anthropicProprietary | 1541 | · | $4 |
| 45 | Gemini 3.6 Flash (High) googleProprietary | 1538 | 69,2 | $1,5 |
| 46 | Claude Opus 4.6 anthropicProprietary | 1537 | · | · |
| 47 | Claude Opus 4.8 anthropicProprietary | 1536 | · | · |
| 48 | Muse Spark 1.2 (xHigh) metaProprietary | 1534 | 72,2 | $2 |
| 49 | Claude Sonnet 4.6 anthropicProprietary | 1523 | · | · |
| 50 | GPT-5.6 Terra (xHigh, Codex Harness) openaiProprietary | 1522 | · | · |
| 51 | Seed2.1 Pro Preview bytedanceProprietary | 1520 | · | · |
| 52 | GPT-5.6 Luna (xHigh, Codex Harness) openaiProprietary | 1519 | · | · |
| 53 | Qwen3.7-Max (20260517) alibabaProprietary | 1515 | · | · |
| 54 | GPT-5.5 (xHigh, Codex Harness) openaiProprietary | 1513 | · | · |
| 55 | Muse Spark Open (Max) metamuse-model-license | 1511 | · | · |
| 56 | Kimi K2.6 moonshotModified MIT | 1510 | · | · |
| 57 | GLM-5.1 zaiMIT | 1508 | · | · |
| 58 | Hy3 tencentApache 2.0 | 1506 | 58,8 | $0,24 |
| 59 | Gemini 3.5 Flash (High) googleProprietary | 1500 | 70,1 | $3,38 |
| 60 | Claude Opus 4.5 (High, 32K, 20251101) anthropicProprietary | 1494 | · | · |
| 61 | Gemini 3.5 Flash (Medium) googleProprietary | 1491 | · | $3,38 |
| 62 | GPT-5.5 (High, Codex Harness) openaiProprietary | 1487 | · | · |
| 63 | Qwen3.6-Max-Preview alibabaProprietary | 1483 | · | $0 |
| 64 | MiniMax-M3 minimaxMiniMax Community License | 1482 | 58,6 | $0,53 |
| 65 | MiMo-V2.5-Pro xiaomiMIT | 1481 | 60,2 | $0,54 |
| 66 | Kimi K2.7 Code moonshotModified MIT | 1474 | 60,8 | $1,71 |
| 67 | Claude Opus 4.5 (20251101) anthropicProprietary | 1469 | · | · |
| 68 | DeepSeek-V4-Pro Preview (High) deepseekMIT | 1464 | · | · |
| 69 | GPT-5.4 (High, Codex Harness) openaiProprietary | 1464 | · | · |
| 70 | Qwen3.6-Plus alibabaProprietary | 1462 | 54,5 | $1,13 |
| 71 | GPT-5.5 (Codex Harness) openaiProprietary | 1456 | · | · |
| 72 | DeepSeek-V4-Pro deepseekMIT | 1447 | · | · |
| 73 | Gemini 3.1 Pro Preview googleProprietary | 1447 | 68,8 | $4,5 |
| 74 | GPT-5.4 (Medium, Codex Harness) openaiProprietary | 1444 | · | · |
| 75 | Gemini 3.5 Flash-Lite googleProprietary | 1441 | 49,3 | $0,85 |
| 76 | Gemini 3 Pro googleProprietary | 1440 | · | · |
| 77 | Gemini 3 Flash googleProprietary | 1440 | · | · |
| 78 | MiMo-V2.5 xiaomiMIT | 1438 | 56,8 | $0,18 |
| 79 | beam-1.0-501b-a23b reflectionApache 2.0 | 1438 | · | · |
| 80 | Kimi K2.5 (Thinking) moonshotModified MIT | 1437 | · | · |
| 81 | GLM-5 zaiMIT | 1435 | · | · |
| 82 | GLM-4.7 zaiMIT | 1435 | · | · |
| 83 | MiMo-V2-Pro xiaomiProprietary | 1434 | · | $0 |
| 84 | DeepSeek-V4-Flash Preview (High) deepseekMIT | 1431 | · | · |
| 85 | GPT-5 (Medium) openaiProprietary | 1418 | · | $3,44 |
| 86 | GPT-5.2 openaiProprietary | 1417 | · | · |
| 87 | Inkling thinkyApache 2.0 | 1414 | 52,1 | $1,76 |
| 88 | Inkling-Small thinkyApache 2.0 | 1409 | 52,9 | $0,53 |
| 89 | Kimi K2.5 (Instant) moonshotModified MIT | 1405 | · | · |
| 90 | GPT-5.4 openaiProprietary | 1402 | · | · |
| 91 | GLM-5V-Turbo zaiProprietary | 1402 | · | · |
| 92 | Qwen3.5-397B-A17B alibabaApache 2.0 | 1400 | · | · |
| 93 | MiniMax-M2.7 minimaxModified MIT | 1399 | 52,6 | $0,53 |
| 94 | GPT-5.4 mini (High) openaiProprietary | 1398 | · | · |
| 95 | GPT-5.1 (Medium) openaiProprietary | 1396 | · | · |
| 96 | Claude Sonnet 4.5 (High, 32K, 20250929) anthropicProprietary | 1394 | · | · |
| 97 | Claude Opus 4.1 (20250805) anthropicProprietary | 1390 | · | · |
| 98 | MiniMax-M2.5 minimaxModified MIT | 1386 | · | $0,53 |
| 99 | Claude Sonnet 4.5 (20250929) anthropicProprietary | 1385 | · | · |
| 100 | Gemini 3 Flash (Minimal) googleProprietary | 1385 | · | · |
Calidad frente a precio
La línea verde une los modelos que nadie supera a la vez en WebDev Arena y en precio: si un modelo queda por debajo, existe otro igual de barato que rinde más. Azul: pesos abiertos. Rojo: propietarios.