Vascend
HorizonVascend IDEVascend CLICommunityBenchmarkContacto
AccederÚnete a Discord
InicioHorizonIDECLIAcceder
ChangelogInteligenciaCoste por tareaCoding Agent IndexCyber IndexDecision IndexBenchmarksVelocidadLatenciaPrecioClasificación
Benchmark

Ministral 3 14B

  • Mistral
  • Lanzamiento · 2 de diciembre de 2025
  • Pesos abiertos
  • Sin razonamiento

Inteligencia

Intelligence Index

#231 de 266
  1. 6,36,3Phi-4 Mini Instruct
  2. 6,26,2Qwen3.5 2B (Non-reasoning)
  3. 6,16,1DeepHermes 3 - Mistral 24B Preview (Non-reasoning)
  4. 6,16,1Jamba 1.7 Large
  5. 6,16,1Qwen3.5 0.8B (Reasoning)
  6. 66Ministral 3 14B
  7. 66Qwen3 Omni 30B A3B Instruct
  8. 66Granite 4.0 H Small
  9. 5,95,9Nova Micro
  10. 5,95,9Phi-4
Datos de Artificial Analysis

Evaluaciones

  • AA-Omniscience Accuracy13,6%#193
  • Humanity's Last Exam4,6%#224
  • Terminal-Bench 4.00%#96
  • AA-LCR26,3%#180
  • SciCode23,8%#132
  • CritPt0%#143
  • MMMU-Pro49,8%#103
  • GPQA Diamond57,2%#150
  • IFBench32%#118
  • τ²-Bench Telecom27,2%#86
  • τ²-Bench Banking6,6%#78
  • Terminal-Bench Hard4,6%#81
Ministral 3 14BMediana
Datos de Artificial Analysis

Velocidad y precio

Velocidad

tokens por segundo · #111 de 162
  1. 8181Gemini 4 Argon (High)
  2. 8181Hy3
  3. 8080Claude Opus 5.5 (High)
  4. 8080Qwen3.5 9B (Non-reasoning)
  5. 7878Grok 4.7 (Low)
  6. 7777Ministral 3 14B
  7. 7777Magistral Small 1.2
  8. 7676LongCat 2.0
  9. 7070GLM-5.3-Flash
  10. 7070Claude Fable 5.1 (Max)
Datos de Artificial Analysis

Latencia

segundos al primer token · #38 de 162
  1. 0,90,9Qwen3 Coder Next
  2. 0,90,9Mistral Small 4 (Non-reasoning)
  3. 11Gemma 4 26B A4B (Non-reasoning)
  4. 11gpt-oss-20b (Low)
  5. 11NVIDIA Nemotron 3 Nano 30B A3B (Reasoning)
  6. 11Ministral 3 14B
  7. 11Claude Sonnet 5.5 (Low)
  8. 11Llama 4 Maverick
  9. 11Nova 2.0 Pro Preview (Non-reasoning)
  10. 11Nova 2.0 Lite (Non-reasoning)
Datos de Artificial Analysis

Precio

$ por 1M tokens (3:1) · #48 de 169
  1. $0,20,2GPT-6 Luna (High)
  2. $0,20,2GPT-6 Luna (Medium)
  3. $0,20,2Claude Haiku 5.5 (Low)
  4. $0,20,2GPT-6 Luna (Low)
  5. $0,20,2GPT-6 Luna (Non-reasoning)
  6. $0,20,2Ministral 3 14B
  7. $0,220,22Phi-4
  8. $0,220,22Gemma 4 31B (Non-reasoning)
  9. $0,230,23Qwen3.8-Flash-Next
  10. $0,240,24GLM-5.3-Flash
Datos de Artificial Analysis

Modelos cercanos

  • Phi-4 Mini InstructMicrosoft6,3
  • Qwen3.5 2B (Non-reasoning)Alibaba6,2
  • DeepHermes 3 - Mistral 24B Preview (Non-reasoning)Nous Research6,1
  • Jamba 1.7 LargeAI21 Labs6,1
  • Qwen3.5 0.8B (Reasoning)Alibaba6,1
  • Qwen3 Omni 30B A3B InstructAlibaba6
  • Granite 4.0 H SmallIBM6
  • Nova MicroAmazon5,9
Datos de Artificial Analysis

Mismo laboratorio

  • Mistral Large 4 PreviewMistral38,4
  • Mistral Medium 3.5Mistral14,2
  • Magistral Medium 1.2Mistral11,8
  • Mistral Small 4 (Reasoning)Mistral11,3
  • Mistral Small 4 (Non-reasoning)Mistral9
  • Magistral Small 1.2Mistral8,6
  • Ministral 3 8BMistral5,5
  • Ministral 3 3BMistral4,8
Datos de Artificial Analysis
HorizonVascend IDEOlympusTalosCursosCommunityBenchmark
DiscordGitHubXHablemosAvisos legalesPrivacidadCookiesUso de datosSeguridad
© 2026 Vascend Technologies · VAT IT04911550616
ascend