DeepSeek: DeepSeek V3.1

Fiabilité 20%

deepseek/deepseek-chat-v3.1

DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates. It extends the DeepSeek-V3 base with a two-phase long-context training process, reaching up to 128K tokens, and uses FP8 microscaling for efficient i...

🧠 Intelligence & Données
Knowledge Cutoff: 2025-03-31
Tokenizer: DeepSeek
Moderation: ✅ Non
📅 Cycle de vie
Ajouté le: 21/08/2025
Spécifications
  • Provider & Modalité deepseek text->text
  • Fenêtre de contexte 32,768 tokens
  • Max Output Tokens 7,168
  • Support des Outils (Tools) ✔️ Fonction Calling
🔍 Modèles similaires
Modèle Provider Input Output Contexte
Reka Edge
rekaai/reka-edge
rekaai $0.1000 $0.1000 16,384
LiquidAI: LFM2-24B-A2B
liquid/lfm-2-24b-a2b
liquid $0.0300 $0.1200 32,768
LiquidAI: LFM2.5-1.2B-Thinking (free)
liquid/lfm-2.5-1.2b-think...
liquid $0.0000 $0.0000 32,768
LiquidAI: LFM2.5-1.2B-Instruct (free)
liquid/lfm-2.5-1.2b-instr...
liquid $0.0000 $0.0000 32,768
Mistral: Mistral Small Creative
mistralai/mistral-small-c...
mistralai $0.1000 $0.3000 32,768