Xiaomi Tech

Let everyone in the world enjoy a better life through innovative technology.

[Xiaomi Tech @XiaomiTech_]

🎬🔊 Meet ControlFoley, Xiaomi's open-source Video-to-Audio model.

Go beyond matching sound to visuals. Use text prompts or reference audio to shape the sound you want, synced with the video.

Try it 👇
https://t.co/6cEaknkuFv

#Xiaomi #AI #OpenSource #AIGC #Multimodal

[Xiaomi Tech @XiaomiTech_]

Our MiMo team just made it happen 🚀 1000+ tokens/s on a 1T parameter model. No custom silicon. No wafer-scale integration. Just a standard 8-GPU node. FP4 quantization + DFlash speculative decoding on the model side, TileRT ultra-low-latency inference on the system side — the result of pure co-design 💪

[Xiaomi MiMo @XiaomiMiMo]

🚀 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: [email protected]

[Xiaomi Tech @XiaomiTech_]

Xiaomi Robotics achieved a dual-champion in embodied AI.

🏆 1st Place at CVPR 2026 Workshops-GigaBrain Challenge-RoboChallenge Track — the only team to achieve a success rate above 40% (40.89%).

🏆 1st Place at ICRA 2026 WBC — achieving a 94% success rate and a score of 99.2.

Bringing embodied AI into the real world, enabling robots to perform useful tasks with greater intelligence and reliability.

#EmbodiedAI #Robotics #VLM #WorldModel #Sim2Real #CVPR2026 #ICRA2026

[Xiaomi Tech @XiaomiTech_]

How did we achieve up to 99% price cut?
Xiaomi MiMo-V2.5 inference optimizations, now public for the first time.
Read more ⬇️

[Xiaomi MiMo @XiaomiMiMo]

What’s new with MiMo-V2.5 series inference?

We just published a blog on our full pipeline inference optimizations for MiMo-V2.5 series, including how we pushed hybrid SWA efficiency to the limit.

Read the full blog here:
https://t.co/lYBEcgaVhU

[Xiaomi Tech @XiaomiTech_]

13 partners 🤝 Integrated into real products and workflows ⚡Xiaomi MiMo Orbit Program is building an open Agent ecosystem — and this is just the beginning 🚀

[Xiaomi MiMo @XiaomiMiMo]

Today we're excited to announce the first 13 ecosystem partners joining the Xiaomi MiMo Orbit Program.

A sincere thank you to every partner for your trust and collaboration. @gitlawb and 12 other ecosystem partners.

This is just the beginning. MiMo Orbit Program is still open, and we'd love to collaborate with you!
→ [email protected]

[Xiaomi Tech @XiaomiTech_]

Xiaomi MiMo-V2.5: lower prices, more tokens, simpler rules. Let's go! 🚀

[Xiaomi MiMo @XiaomiMiMo]

🚀 Better inference efficiency, lower costs, broader access.

MiMo-V2.5 Series API pricing is now permanently reduced — by up to 99% compared to previous pricing.
✨ Unified pricing across all context lengths.
MiMo Token Plans have also been upgraded:
• 5–8× more usable tokens at the same price
• Simpler and more transparent billing rules
🎁 As a thank-you to current users, all current Token Plan credits will be fully reset.
🎧 MiMo-V2.5-TTS remains free for a limited time.
⏰ Effective May 26 at 6:00 PM PDT.

These improvements are powered by continued inference optimization and serving efficiency upgrades across the MiMo stack.
🛠️ We’ll also publish a detailed technical blog on the inference optimizations later — stay tuned.

[Xiaomi Tech @XiaomiTech_]

For years, 3D reconstruction and video generation in assisted driving have been separate pipelines. Xiaomi Auto World Model's brand-new framework changes that — deeply coupling both in a single architecture, achieving SOTA on Waymo & nuScenes. Learn more 👇

🔗 https://t.co/IBM8cZC36b

#AssistedDriving #WorldModel #3DReconstruction #VideoGeneration #ComputerVision #XiaomiAuto #XiaomiSU7

[Xiaomi @Xiaomi]

Join us for #XiaomiLaunch on May 28th at 14:00 (GMT+2). 🟠📷🔴
Beyond sight, beyond now. https://t.co/4TZjUpzpKE

[Xiaomi Tech @XiaomiTech_]

Reflection Removal 🥈 — distilled a diffusion model's generative prior into a fast discriminative network. Strong + fast.📎 https://t.co/pbwR0mxujY

[Xiaomi Tech @XiaomiTech_]

🏆 Xiaomi's imaging algorithms team won 3 awards at CVPR 2026 NTIRE — 1st in Efficient SR, 1st in Portrait Restoration, 2nd in Reflection Removal.🧵 A quick look at each👇

#CVPR2026 #NTIRE #ComputerVision #XiaomiTech

[Xiaomi Tech @XiaomiTech_]

Portrait Restoration 🥇 — two-stage cascade: "fix structure → details → aesthetics", with one-step diffusion replacing slow multi-step sampling.📎https://t.co/TOMNF9QBdz

[Xiaomi MiMo @XiaomiMiMo]

📢Calling all Apache Software Foundation committers

Xiaomi MiMo is giving you our Max Token Plan for FREE as part of the 100T Token Grant for Builders Program.
Sign up with your https://t.co/PpNIfNBsX7 email → instantly activated.
Sign up now: https://t.co/KmdO2TIxKq

[Xiaomi @Xiaomi]

The wait is almost over. It's finally happening.

Xiaomi 17T Series launches May 28th.

Mark your calendars for our biggest T Series upgrade yet!