Frame-X.AI

[Frame-X.AI @FrameX_ai]

Introducing Rise — the Omni-Immersive Foundation Model in the Era series.

Your camera is the controller.
You turn. The world turns.
Under 500ms, end to end.

Try it (free this week):
https://t.co/m5DhSrVFo1
Show us what you steer with.

[Frame-X.AI @FrameX_ai]

For 70 years, hardware defined how we entered worlds. Controllers, wheels, headsets.

Rise is our first step past that.
From experiencing worlds to entering them.

Try it (free this week): https://t.co/m5DhSrVFo1
Show us what you steer with.

[Frame-X.AI @FrameX_ai]

Introducing Rise — the Omni-Immersive Foundation Model in the Era series.

Your camera is the controller.
You turn. The world turns.
Under 500ms, end to end.

Try it (free this week):
https://t.co/m5DhSrVFo1
Show us what you steer with.

[Frame-X.AI @FrameX_ai]

That's the point. The model isn't learning "this gesture → that output." It's learning how people, environments, and generated worlds move together.

We call it Native Interactive Representation Scaling: interaction scales *with* the model, not around it.

[Frame-X.AI @FrameX_ai]

Hi Ivan — our CEO would love to personally host you for a quick hands-on demo of Genesis, our real-time interactive video model. His team was recognized among China’s Top 6 foundation-model teams, alongside DeepSeek, Qwen and Kimi. We’d love to welcome you while you’re in Beijing!

[Frame-X.AI @FrameX_ai]

Hi Ivan — your China itinerary looks fantastic. Great to see Beijing as your first stop, and we’d love to host you while you’re in town.

I’m reaching out from Frame-X. We recently launched Genesis, our first interactive video-generation model under the Era series. It’s a full-duplex system that processes live camera input, allows natural conversation interruptions, and delivers real-time responses.

Our core team, led by our CEO, was named one of China’s Top 6 foundation-model teams of 2025 by AI researcher Simon Willison, together with DeepSeek, MiniMax, Zhipu AI, Qwen and Kimi.

If your timeline permits, we’d welcome you for a brief hands-on demo to exchange thoughts. We can work around your schedule — feel free to DM us. Safe travels, and looking forward to meeting you in Beijing!

[Frame-X.AI @FrameX_ai]

Hi, Genesis.

[Frame-X.AI @FrameX_ai]

@_akhaliq Thanks @_akhaliq for sharing our work!
For the first time in industry, stream-based video model outperforms conventional non-stream ones across multiple dimensions (quality, motion smoothness, etc) simultaneously.
Code: https://t.co/WTYibQERDD
Project: https://t.co/fzKrfeAuhh

[Frame-X.AI @FrameX_ai]

Thanks @_akhaliq for sharing our work!
For the first time in industry, stream-based video model outperforms conventional non-stream ones across multiple dimensions (quality, motion smoothness, etc) simultaneously.
Code: https://t.co/WTYibQERDD
Project: https://t.co/fzKrfeAuhh

[AK @_akhaliq]

Stream-R1

Reliability-Perplexity Aware Reward Distillation for Streaming Video Generation

paper: https://t.co/Q8vlxQcUyP