正在加载视频...

视频加载失败

One image + text + camera trajectory = controllable worlds. All on a single GPU. Our research team just released SANA-WM, a 2.6B open source world model natively trained for 60-second video generation with precise camera control.

93,496 次观看 • 6 个月前 •via X (Twitter)

33 条评论

NVIDIA AI 的头像
NVIDIA AI6 个月前

The architecture features Hybrid Linear Attention, Dual-Branch Camera Control, Two-Stage Generation Pipeline, and a Robust Annotation Pipeline. All of this combines for stronger action-following accuracy and higher throughput while maintaining visual quality.

NVIDIA AI 的头像
NVIDIA AI6 个月前

You can find the full project including the paper, code, and model here 👇

PublicAI 的头像
PublicAI6 个月前

this sounds super cool! creating worlds with just a single gpu is a game changer. can't wait to see what people do with SANA-WM!

The AI Colony R&D 的头像
The AI Colony R&D5 个月前

60 seconds of video. One image to start. One GPU to run it. Open source. The game engine just became a prompt.

PaperTrace 的头像
PaperTrace6 个月前

Looks great. Can you link the paper and repo, name training data sources/licenses, and state the exact GPU/VRAM for the single GPU claim? That would let others reproduce the camera control results.

Kekko D’Amato 的头像
Kekko D’Amato6 个月前

2.6B for 60-second world-consistent video with camera control is a wild benchmark. The open-source release is what makes this actually useful — most teams don't need Hollywood quality, they need something they can run and fine-tune. What hardware are people actually running inference on?

Naveen 的头像
Naveen5 个月前

love how you're pushing the boundaries of video generation with SANA-WM

Supersocks 的头像
Supersocks6 个月前

@llmgram

Jonathan 🇺🇲 的头像
Jonathan 🇺🇲6 个月前

You're on the right track Nvidia, but please make it more performant and keep training. Go full send on compute for training.

Pruthviraj P 的头像
Pruthviraj P5 个月前

SANA-WM 2.6B open source world model for 60-second video on a single GPU

Creative Side-B 的头像
Creative Side-B5 个月前

Looking very soulless, generic and not worth it!

Ajeya 的头像
Ajeya5 个月前

Can it run on a 4090?

Dray 的头像
Dray6 个月前

$NVDA seriously is going to rule the world

Apex 的头像
Apex6 个月前

One image + text + trajectory = controllable worlds on a single GPU. Simulation is becoming cheap and precise.

Chiko Chiko 的头像
Chiko Chiko6 个月前

bbox queen looks clean, seedance 2 with gpt image 2, huh? got a side by side without seedance to compare?

Tim 的头像
Tim6 个月前

Love this

Asim Hussain 的头像
Asim Hussain6 个月前

"single GPU" is the part that changes everything here. world models for video have always felt like a data center problem, shrinking that down to one GPU opens this up to a completely different audience

XtaxRich | AI Tool Tests 的头像
XtaxRich | AI Tool Tests5 个月前

SANA-WM is concrete enough to test. One image, text, camera path, and single GPU are the setup. Control drift is the receipt.

舔狗AI实验室 🛠️🐶 的头像
舔狗AI实验室 🛠️🐶5 个月前

2.6B for 60-second generation is compelling if SANA-WM keeps temporal drift under control.

Eco 的头像
Eco5 个月前

@KashJim50723 🤷‍♂️

OctalSun 的头像
OctalSun5 个月前

Which GPU, and how much VRAM?

Mattias Johansson 的头像
Mattias Johansson5 个月前

Hey @NVIDIAAIDev — got SANA-WM running on one H100. "Single GPU" holds at runtime, but the install path is rough: HF repo has no model_index.json, real setup needs vendoring the feat/sana-wm PR branch + mmcv 1.7.2 + flash-attn stub. Any chance of an official quickstart? 🙏

Yon Learning 的头像
Yon Learning6 个月前

What's the max resolution you folks are targeting for high-res video?

Aroma Aura 的头像
Aroma Aura5 个月前

Awesome share

TechHalla 的头像
TechHalla5 个月前

Oh I wanna test this

AI Mastery Guide 的头像
AI Mastery Guide6 个月前

One image plus text plus camera trajectory and you get a controllable world. That formula is deceptively simple for what it produces

Lucky harry 的头像
Lucky harry5 个月前

wao great now give me 100 B200 for my startup.

Capybara 的头像
Capybara6 个月前

Wuuut 👀😳🤯

Synthella 的头像
Synthella5 个月前

Interesting how "video generation" is quietly evolving into controllable spatial simulation.

ArtiMindArt 的头像
ArtiMindArt5 个月前

thats awesome! pretty cool!

Gregor 的头像
Gregor6 个月前

what's the latency on a consumer GPU though?

Govin.AI 的头像
Govin.AI5 个月前

Finally a world model that doesn’t require a datacenter to test your camera moves

Rifat Ahmed 的头像
Rifat Ahmed6 个月前

means i have to huy it ASAP🥹

相关视频

NVIDIA just unleashed SANA-WM and it’s an absolute MONSTER for the future of open source AI! A blazing-fast 2.6B-parameter open-source world model that doesn’t just generate video… it creates controllable, physics-rich, high-fidelity worlds on demand. Why this is insanely powerful: • One image + text prompt + 6-DoF camera trajectory → generates 720p videos up to 60 seconds long with buttery-smooth, precisely controlled camera movement. You’re not just watching, you’re piloting the simulation. • Runs locally on a single consumer GPU (RTX 5090 level) thanks to heavy distillation + NVFP4 quantization. Full 60-second clip denoised in ~34 seconds. No massive clusters required. • 36× higher throughput than previous open models while rivaling (or beating) closed industrial giants in visual quality and consistency. • Trained lightning-fast: ~213K public videos in just 15 days on 64 H100s. • Built with next-level tech: Hybrid Linear Attention, dual-branch camera control, two-stage pipeline, and rock-solid metric-scale pose understanding. This is a true open world model, the foundation for embodied AI, robotics, autonomous systems, and hyper-realistic simulations that can run anywhere. Project: At our Zero-Human Company, we’re already running SANA-WM live in our core pipelines. It’s supercharging autonomous agent training, generating unlimited synthetic training data, and powering full end-to-end simulation loops, zero humans in the loop. The speed and control let us test thousands of edge-case scenarios overnight, iterate at lightspeed, and push our fully autonomous operations further than ever before. This is the kind of breakthrough that turns science fiction into daily reality. World models just leveled up — hard. The age of personal, local, controllable universes is here.

Brian Roemmele

619,954 次观看 • 6 个月前