Research · Updated 25 Oct, 03:38 pm IST
Nvidia open-sources Flora to label robot video far cheaper than large LLMs
Why it matters for readers: If you follow AI for robots, Flora shows how Nvidia automates a lot of the expensive video labeling that robotic systems need to learn.
- Nvidia says Flora can label one hour of robot video for about $0.72 versus $88 with GPT-6 Astra in its comparison.1
- Flora produces three kinds of labels: action-time segmentation, step-by-step execution ratings, and task-success judgments.1
- Flora is built from three fine-tuned Qwen3.5 models: a 9B model for segmenting actions and two 27B models for describing actions and providing reward/evaluation roles, including Flora-Rewarder.1
- Nvidia reports Flora is running during training of its GR00T models, annotating up to 200,000 hours of video per day across 512 GB300 GPUs, and the models are available on Hugging Face.1
Get a brief like this every morning
Uzha reads hundreds of sources and gives you the stories that matter for your work, with every source linked. Free.
Get started