🚀 1,000+ TOKENS/S ON A 1T MODEL! 🚀
We are thrilled to release Xiaomi MiMo-V2.5-Pro-UltraSpeed in collaboration with @TileRT_AI , breaking the 1,000 tokens/s output speed on a 1 Trillion parameter model for the FIRST TIME!
Not wafer-scale integration like Cerebras. Not pure on-chip SRAM chips like Groq. We achieve 1,000 tps on a 1T MoE model using just a SINGLE, STANDARD 8-GPGPU NODE.
Read the full technical deep dive：https://t.co/MX0kjHKdKi
Want to experience the future of real-time AI?
👉 Apply for UltraSpeed now: https://t.co/aeWAxyhwVk
⏳ Limited-Time Access: Application-based · Jun 8 – Jun 23 (PDT)
💬 Chat Experience: Completely FREE for a limited time — try the blazing-fast web chat now.
⚡ UltraSpeed API: Just 3x the price for a ~10x boost in output experience.
🤝 Enterprise & Large-Scale Needs: business-mimo@xiaomi.com
🔓 And the best part — we're open-sourcing it.
1,000+ tps on a 1T model wasn't a single breakthrough — it's deep model × system co-design between the MiMo and TileRT teams, all on general-purpose GPUs (no Cerebras-style wafer-scale, no Groq-style SRAM ASICs).
On the model side: FP4 quantization (smaller footprint, less memory traffic) + DFlash, our block-masked parallel speculative decoding that accepts far more tokens per verification. On the system side, TileRT tailors its compiler & kernels to exactly these techniques.
The result: a 1T model breaking 1,000 tps on a single, standard 8-GPU node.
🤗 Open weights (FP4 + DFlash checkpoint): https://t.co/jYQsgeruMg
mixture-of-experts — A model built from many specialist sub-networks where only a few activate per token, giving big-model capability at small-model running cost.
quantization — Shrinking a model by storing its numbers less precisely — like rounding — so it runs faster and fits on smaller hardware, at a small quality cost.
speculative decoding — A speed trick where a small model drafts several tokens ahead and the big model verifies them in one pass, often doubling generation speed.
AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
Why it matters
Shows frontier-scale decode speed is reachable on commodity 8-GPU nodes through FP4 plus speculative decodingA speed trick where a small model drafts several tokens ahead and the big model verifies them in one pass, often doubling generation speed.Full definition →, which changes the latency you can assume for large-model AI agentAn AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.Full definition → loops without specialty silicon.
postHeads up, agent users！ If you're using Xiaomi MiMo with thinking mode: When…
postMiMo-V2.5 and V2.5-Pro go open weights under MIT with day-zero SGLang and vLLM
postIntroducing MiMo-V2.5 Voice — our full-stack voice lineup for the Agent era. 🚀…
