API usage

The 20 open models served most on OpenRouter, tokens per day over 30 days.

All of them, tokens per day

Last 7 days 56.2T tokens −5% compared with the 7 days before
  1. 1DeepSeek-V4.1-FlashDeepSeek
    15.9T tokens−6% compared with the 7 days before
  2. 2GLM-5.3-FlashZ.ai
    11.5T tokens+2% compared with the 7 days before
  3. 3GLM-5.3Z.ai
    3.6T tokens−1% compared with the 7 days before
  4. 4NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16NVIDIA
    3.4T tokens−8% compared with the 7 days before
  5. 5Kimi-K3.1-ProMoonshot AI
    3.4T tokens−23% compared with the 7 days before
  6. 6DeepSeek-V4-Flash-0731DeepSeek
    3.1T tokens+1% compared with the 7 days before
  7. 7NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16NVIDIA
    3.1T tokens−4% compared with the 7 days before
  8. 8DeepSeek-V4-FlashDeepSeek
    2T tokens−10% compared with the 7 days before
  9. 9Inkling-2-SmallThinking Machines
    1.6T tokens+1065% compared with the 7 days before
  10. 10Hy3Tencent
    1.4T tokens−36% compared with the 7 days before
  11. 11Kimi-K3Moonshot AI
    1.2T tokens−25% compared with the 7 days before
  12. 12NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16NVIDIA
    1.1T tokens0% compared with the 7 days before
  13. 13GLM-5.2Z.ai
    955.5B tokens+5% compared with the 7 days before
  14. 14MiMo-V2.6-Pro-RLXiaomi MiMo
    911.4B tokens−20% compared with the 7 days before
  15. 15Qwen3.8-27BQwen
    826.8B tokens+12% compared with the 7 days before
  16. 16DeepSeek-V4-ProDeepSeek
    556.2B tokens+1% compared with the 7 days before
  17. 17MiniMax-M3MiniMax
    508.5B tokens−23% compared with the 7 days before
  18. 18gemma-4-26B-A4B-itGoogle
    447.3B tokens+12% compared with the 7 days before
  19. 19gemma-4-31B-itGoogle
    366.2B tokens−19% compared with the 7 days before
  20. 20gpt-oss-120bOpenAI
    80.9B tokens−73% compared with the 7 days before

Source: OpenRouter (openrouter.ai/rankings), as of Nov 17, 2026.