Rankings de modelos
Rankings de modelos
Cada fuente usa su propio criterio. No hay un #1 global. Para caídas, ve a Estado.
Arena score
People pick the better answer in blind A/B chats; Arena publishes the resulting score.
Mira: How chat and writing feel to people.
No mira: Truthfulness, security, price, speed, or coding skill.
Top
Arena score →
- 01Gemini 4 CarbonHighgoogle1539 Arena score
- 02Gemini 4 ArgonHighgoogle1530 Arena score
- 03Claude Fable 5.1Maxanthropic1513 Arena score
- 04Claude Opus 5.5Highanthropic1512 Arena score
- 05Claude Opus 5Maxanthropic1507 Arena score
- 06Muse WonderMaxmeta1505 Arena score
- 07Claude Opus 5Highanthropic1503 Arena score
- 08Claude Opus 4.6Highanthropic1502 Arena score
- 09Gemini 3.8 FlashHighgoogle1501 Arena score
- 10Claude Opus 4.6anthropic1497 Arena score
- 11Claude Fable 5Highanthropic1492 Arena score
- 12MiMo-V2.6-Pro1491 Arena score
- 13Muse Spark 1.3Maxmeta1491 Arena score
- 14Claude Opus 4.7Highanthropic1490 Arena score
- 15Gemini 3.7 FlashHighgoogle1488 Arena score
- 16Qwen3.8-Maxalibaba1484 Arena score
- 17Muse Spark 1.2xHighmeta1484 Arena score
- 18Claude Opus 4.7anthropic1483 Arena score
- 19Gemini 3.5 FlashHighgoogle1482 Arena score
- 20Gemini 3.1 Pro Previewgoogle1480 Arena score
Las barras se escalan para que las diferencias se vean mejor.
El mismo modelo en cada fuente. Vacío = no está en ese top.