Clasificación
Clasificación: Contexto largo
Búsqueda y razonamiento en entradas muy largas: MRCR v2 y ventana de contexto.
General767Código469Matemáticas90Razonamiento483Agéntico89Conocimiento482Multimodal452Agentes de código68Razonamiento científico96Multilingüe12Escritura95Chat y asistentes95Investigación99Llamada de herramientas45Contexto largo116Velocidad472Valor471Pesos abiertos183
116 modelos
| # | Modelo | Podium Score | Velocidad |
|---|---|---|---|
| 01 | Kimi K3.1 Pro Moonshot AIAbiertorazonamiento | 100.0 | 45 t/s |
| 02 | GPT-6 Astra OpenAIrazonamiento | 100.0 | 82 t/s |
| 03 | GPT-5.6 Sol OpenAI | 100.0 | 72 t/s |
| 4 | GPT-5.4 OpenAI | 100.0 | 60 t/s |
| 5 | Claude Fable 5 Anthropicrazonamiento | 99.2 | 71 t/s |
| 6 | Kimi K3 Moonshot AIAbiertorazonamiento | 99.2 | 37 t/s |
| 7 | Claude Opus 5 Anthropic | 99.2 | 60 t/s |
| 8 | Qwen3.8 Max Alibaba | 99.2 | 55 t/s |
| 9 | Claude Opus 4.8 Anthropic | 99.2 | 50 t/s |
| 10 | Muse Spark 1.1 Meta | 99.2 | 208 t/s |
| 11 | Gemini 3.6 Flash Google | 99.2 | 233 t/s |
| 12 | Claude Opus 4.7 Anthropic | 99.2 | 80 t/s |
| 13 | Qwen3.7 Max Alibaba | 99.2 | 203 t/s |
| 14 | Gemini 3.5 Flash Google | 99.2 | 267 t/s |
| 15 | MiMo V2.5 Pro XiaomiAbierto | 99.2 | 65 t/s |
| 16 | DeepSeek V4 Pro DeepSeekAbierto | 99.2 | 66 t/s |
| 17 | Claude Sonnet 5 Anthropic | 99.2 | 96 t/s |
| 18 | Qwen3.7 Plus Alibaba | 99.2 | 54 t/s |
| 19 | DeepSeek V4 Flash DeepSeekAbierto | 99.2 | 113 t/s |
| 20 | Gemini 3.5 Flash-Lite Google | 99.2 | 389 t/s |
| 21 | Claude Sonnet 4.6 Anthropic | 99.2 | 50 t/s |
| 22 | Qwen3.6 Plus Alibaba | 99.2 | 55 t/s |
| 23 | Gemini 4 Argon Googlerazonamientovista previa | 99.0 | 88 t/s |
| 24 | GPT-5.6 Terra OpenAI | 98.4 | 127 t/s |
| 25 | Claude 5.5 Sonnet Anthropicrazonamiento | 98.0 | 94 t/s |
| 26 | Kimi K3.1 Highspeed Moonshot AIAbiertorazonamiento | 98.0 | 130 t/s |
| 27 | DeepSeek V4.1 Flash DeepSeekAbiertorazonamiento | 98.0 | 214 t/s |
| 28 | GPT-6.1 Sol OpenAIrazonamiento | 96.0 | 75 t/s |
| 29 | Gemini 3.1 Pro Preview Googlerazonamientovista previa | 96.0 | 105 t/s |
| 30 | MiMo v2.6 Flash XiaomiAbierto | 95.0 | 118 t/s |
| 31 | Agnes 3.0 Flash Agnesrazonamiento | 95.0 | 235 t/s |
| 32 | Gemini 3.1 Flash Lite Preview Googlevista previa | 94.0 | 210 t/s |
| 33 | Inkling Small Thinking Machines | 93.6 | 130 t/s |
| 34 | MiniMax M3 MiniMax | 93.6 | 93 t/s |
| 35 | Inkling Thinking Machines | 93.6 | 81 t/s |
| 36 | Nemotron 3 Ultra 550B NVIDIAAbierto | 93.4 | 142 t/s |
| 37 | Grok 4.5 xAI | 93.2 | 59 t/s |
| 38 | Grok 4.3 xAI | 93.2 | 129 t/s |
| 39 | GPT-5.3 Codex OpenAI | 91.2 | 118 t/s |
| 40 | GPT-5.1 OpenAI | 91.2 | 60 t/s |
| 41 | GPT-5 OpenAI | 91.2 | 60 t/s |
| 42 | NVIDIA Nemotron 3.5 Super-VL NVIDIAAbierto | 91.0 | 110 t/s |
| 43 | Claude 5.5 Haiku Anthropicrazonamiento | 91.0 | 145 t/s |
| 44 | DeepSeek V4 Flash Vision DeepSeek | 90.0 | 120 t/s |
| 45 | GLM-5.3 Zhipu AI | 90.0 | 90 t/s |
| 46 | LongCat 2.0 LongCatAbierto | 90.0 | 42 t/s |
| 47 | Mistral Large 4 Mistralrazonamiento | 88.0 | 92 t/s |
| 48 | GLM-5.3 Flash Z.aiAbierto | 88.0 | 182 t/s |
| 49 | GLM-5.2 Z.aiAbierto | 87.5 | 164 t/s |
| 50 | Kimi K2.6 Moonshot AIAbierto | 87.5 | 40 t/s |
| 51 | Nex N2 Pro Nex AGI | 87.5 | 138 t/s |
| 52 | Kimi K2.7 Code Moonshot AIAbierto | 87.5 | 39 t/s |
| 53 | Hunyuan Hy3 Tencent | 87.5 | 69 t/s |
| 54 | Gemma 4 31B GoogleAbierto | 87.5 | 35 t/s |
| 55 | Qwen3.5 397B-A17B AlibabaAbierto | 87.5 | 71 t/s |
| 56 | Qwen3.5 122B-A10B AlibabaAbierto | 87.5 | 126 t/s |
| 57 | Qwen3 Max AlibabaAbierto | 87.5 | 85 t/s |
| 58 | Claude Opus 4.6 Anthropic | 86.3 | 50 t/s |
| 59 | Qwen3.8 Flash Next AlibabaAbierto | 86.0 | 168 t/s |
| 60 | Claude Opus 4.5 Anthropic | 85.2 | 50 t/s |
| 61 | Claude Sonnet 4.5 Anthropic | 85.2 | 80 t/s |
| 62 | o3 OpenAIrazonamiento | 85.2 | 90 t/s |
| 63 | Claude Opus 4.1 Anthropic | 85.2 | 80 t/s |
| 64 | GPT-5.5 OpenAI | 84.9 | 60 t/s |
| 65 | DeepSeek V3.2 DeepSeekAbierto | 83.5 | 60 t/s |
| 66 | DeepSeek R1 DeepSeekAbiertorazonamiento | 83.5 | 60 t/s |
| 67 | GPT-5.2 OpenAI | 81.3 | 90 t/s |
| 68 | Simba 3.2 Independentrazonamiento | 75.0 | 62 t/s |
| 69 | 75.0 | 59 t/s | |
| 70 | 75.0 | 59 t/s | |
| 71 | 75.0 | 62 t/s | |
| 72 | KAT-Coder-Pro V2 Kwai | 75.0 | 103 t/s |
| 73 | Qwen3.8 27B (xhigh) AlibabaAbierto | 75.0 | 53 t/s |
| 74 | Qwen3.8 27B (low) AlibabaAbierto | 75.0 | 64 t/s |
| 75 | Qwen3.8 27B (medium) AlibabaAbierto | 75.0 | 62 t/s |
| 76 | Qwen3.8 27B (Non-reasoning) AlibabaAbiertorazonamiento | 75.0 | 62 t/s |
| 77 | GLM-4.5 Air Zhipu AIAbiertorazonamiento | 72.0 | 96 t/s |
| 78 | Granite 4.2 30B Instruct IBMAbierto | 70.0 | 78 t/s |
| 79 | Ling 3.1 Flash Ling | 68.0 | 175 t/s |
| 80 | Granite 4.2 8B Instruct IBMAbierto | 68.0 | 115 t/s |
| 81 | Granite 4.2 3B Instruct IBMAbierto | 65.0 | 148 t/s |
| 82 | FLUX.2 Pro Black Forest Labs | 62.0 | 30 t/s |
| 83 | Agnes 2.5 Pro Beta Agnes | 62.0 | 65 t/s |
| 84 | GPT-Image 2 OpenAI | 60.0 | 25 t/s |
| 85 | FLUX.2 Max Independent | 60.0 | 35 t/s |
| 86 | GPT-5.6 Luna OpenAI | 56.6 | 175 t/s |
| 87 | 55.0 | 34 t/s | |
| 88 | FLUX.2 Dev Black Forest LabsAbierto | 55.0 | 28 t/s |
| 89 | ElevenLabs Scribe v2 ElevenLabs | 50.0 | 240 t/s |
| 90 | Kling 3.0 Pro Kwai | 50.0 | 25 t/s |
| 91 | G9v3-39A5B AI9StarsAbierto | 50.0 | 60 t/s |
| 92 | G9v3-3B AI9StarsAbierto | 50.0 | 60 t/s |
| 93 | HyperNova 60B MultiverseAbierto | 50.0 | 367 t/s |
| 94 | LFM2.5-2.6B Liquid AIAbierto | 50.0 | 60 t/s |
| 95 | LFM2.5-8B-A1B Liquid AIAbierto | 50.0 | 338 t/s |
| 96 | LFM2.5-VL-1.6B Liquid AIAbierto | 50.0 | 395 t/s |
| 97 | Magistral Small 1.2 MistralAbierto | 50.0 | 60 t/s |
| 98 | Voxtral 3 TTS Mistral | 48.0 | 165 t/s |
| 99 | Eleven Multilingual v3 ElevenLabs | 45.0 | 200 t/s |
| 100 | Cartesia Ink Whisper Cartesia | 45.0 | 260 t/s |
Podium Score es una mezcla ponderada de cinco clasificaciones. Consulta la Metodología para más detalles.