Cheapest LLM APIs (output price ranking)

1. 🥇
   Qwen-Turbo通义千问 Qwen
   $0.20/1M output
2. 🥈
   MiMo V2.5小米 MiMo

   cache hit $0.003/1M
   $0.28/1M output
3. 🥉
   DeepSeek V4 FlashDeepSeek

   cache hit $0.003/1M
   $0.28/1M output
4. 4
   GPT-4o miniOpenAI

   cache hit $0.075/1M
   $0.60/1M output
5. 5
   MiMo V2.5 Pro小米 MiMo

   cache hit $0.003/1M
   $0.83/1M output
6. 6
   DeepSeek V4 ProDeepSeek

   cache hit $0.004/1M
   $0.87/1M output
7. 7
   Doubao Seed 1.6字节 豆包
   $1.1/1M output
8. 8
   Step 3.7 Flash阶跃 StepFun

   cache hit $0.038/1M
   $1.1/1M output
9. 9
   MiniMax-M3MiniMax

   cache hit $0.058/1M
   $1.2/1M output
10. 10
    GPT-4.1 miniOpenAI

    cache hit $0.10/1M
    $1.6/1M output
11. 11
    Qwen-Plus通义千问 Qwen
    $2.4/1M output
12. 12
    Grok 4.3xAI

    cache hit $0.20/1M
    $2.5/1M output
13. 13
    GLM-5.1智谱 Zhipu

    cache hit $0.18/1M
    $3.3/1M output
14. 14
    Kimi K2.6Moonshot Kimi

    cache hit $0.15/1M
    $3.8/1M output
15. 15
    Kimi K2.7 CodeMoonshot Kimi

    cache hit $0.18/1M
    $3.8/1M output

<iframe src="https://www.tokrace.com/embed/best/cheapest?lang=en" width="400" height="360" style="border:0;border-radius:12px" loading="lazy" title="Cheapest LLM APIs (output price ranking)"></iframe>

· Speed is the median of anonymous benchmarks; prices follow official pages and cross-currency uses a fixed USD rate. Reference only.