Fastest DeepSeek Models

AI models ranked by output speed (tokens per second, p50). The fastest large language models for low-latency and high-throughput applications.

1. 1D
   deepseek-chat-v3.1

   ReasoningToolsJSON28.1 intel · $0.210/M · 330ms ttft
   92 t/s

   Speed
2. 2D
   deepseek-v4-flash

   ReasoningToolsJSON46.0 intel · $0.090/M · 537ms ttft
   73 t/s

   Speed
3. 3D
   deepseek-r1

   ReasoningToolsJSON18.8 intel · $0.700/M · 1.4s ttft
   73 t/s

   Speed
4. 4D
   deepseek-v3.2

   ReasoningToolsJSON41.7 intel · $0.229/M · 534ms ttft
   59 t/s

   Speed
5. 5D
   deepseek-v4-pro

   ReasoningToolsJSON39.3 intel · $0.435/M · 664ms ttft
   58 t/s

   Speed
6. 6D
   deepseek-chat-v3-0324

   ToolsJSON22.3 intel · $0.200/M · 846ms ttft
   36 t/s

   Speed
7. 7D
   deepseek-r1-distill-llama-70b

   Reasoning$0.800/M · 761ms ttft · 128K ctx
   31 t/s

   Speed
8. 8D
   deepseek-r1-0528

   ReasoningToolsJSON27.1 intel · $0.500/M · 695ms ttft
   30 t/s

   Speed
9. 9D
   deepseek-v3.1-terminus

   ReasoningToolsJSON28.5 intel · $0.270/M · 881ms ttft
   27 t/s

   Speed
10. 10D
    deepseek-r1-distill-qwen-32b

    ReasoningJSON$0.290/M · 859ms ttft · 128K ctx
    23 t/s

    Speed
11. 11D
    deepseek-chat

    ToolsJSON$0.200/M · 560ms ttft · 131K ctx
    22 t/s

    Speed
12. 12D
    deepseek-v3.2-exp

    ReasoningToolsJSON32.1 intel · $0.270/M · 1.5s ttft
    19 t/s

    Speed