Xiaomi MiMo is 15x Faster Than ChatGPT and Claude

Xiaomi has introduced the MiMo-V2.5-Pro-UltraSpeed mode for its 1-trillion-parameter model, delivering speeds of up to 1,200 tokens per second on standard hardware.

Compiled by Sergey KostenchukPublished 2026-06-11Updated 2026-06-11

2026-06-11 Research OpenAI

🚀 Xiaomi MiMo is 15x Faster Than ChatGPT and Claude

Xiaomi has released the MiMo-V2.5-Pro-UltraSpeed mode for its 1-trillion-parameter model. By utilizing FP4 quantization methods and DFlash speculative decoding, speeds have reached 1,000–1,200 tokens per second on a standard node with 8 GPUs.

🌍 This breakthrough in inference efficiency on standard hardware reduces dependency on proprietary accelerators.

👤 The technology enables the use of powerful models with near-zero latency.

Source 1: https://decrypt.co/370449/xiaomi-mimo-ultraspeed-ai-model-faster-chatgpt-claude

Sources