From the NVIDIA edition of October 25, 2026
NVIDIA GEAR Lab releases Flora, an open model suite for labelling robot and human video, claiming ~125x lower cost than GPT-6 Astra
NVIDIA Research (GEAR Lab) · Research enterprise (robotics and embodied AI) · Global (open models on Hugging Face)
- Publisher
- NVIDIA Research
- Original publication
- Undated project page (paper 2026); covered by The Decoder on 2026-10-25
- Source retrieved
- 2026-10-26
What happened
NVIDIA's GEAR Lab describes Flora as an industrial-strength data engine used to train its GR00T robot models, producing timestamped captions, process-reward and outcome-reward labels from video. NVIDIA says it reaches state-of-the-art scores on WGO-Bench, RoboReward and Robometer, is competitive with GPT-6 Astra on internal benchmarks, is roughly 125x cheaper than API-based alternatives, and labels 200,000 hours of data per day across 512 GB300 GPUs. Its own benchmark table shows Flora trailing GPT-6 Astra on several rows, including the Robot200 DenseCaptionBench subset and WGO-Bench gold-window accuracy, while leading on failure recall (53.3 versus 0.0).
Why it matters
Limited SLED relevance: matters mainly to university robotics and AI research groups evaluating open annotation tools. No public-sector use is described.
Evidence and measured results
Vendor research page and paper with benchmark tables; cost comparison is NVIDIA's estimate. Independent reporting in this edition notes the cost figures are projections.
Limitations and uncertainty
Vendor benchmarks; cost estimates not production measurements; page undated.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-10-26; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- Robotics labs need large volumes of labelled video but commercial annotation APIs are expensive.
- Stakeholders
- University robotics or AI lab directors
- Research computing leads
- Technology transfer / licensing office
- Sponsored research office
- Discovery questions
- Does the lab label robot or human-demonstration video at scale today, and at what cost?
- Will labels or models feed any commercial or sponsored work?
- What GPU capacity is available for batch annotation?
- Value hypothesis
- Interpretation: open annotation models may lower labelling costs for academic robotics labs; savings are unmeasured outside NVIDIA's estimates.
- Potential engagement
- Licensing review and small benchmark of Flora against the lab's current labelling process.
- Unsupported claims
- That labs will see 125x savings
- That Flora beats GPT-6 Astra on robot video
- That commercial use is allowed
Pre-sales engineering
Role takeaway
- Fit
- Academic robotics and embodied-AI research groups with GPU access; weak fit for most state and local government.
- Architecture and integration
- Three fine-tuned Qwen3.5 models (9B segmenter, 27B captioner, 27B rewarder) run as batch jobs on research GPUs over video datasets.
- Prerequisites
- Multi-GPU capacity for 27B models
- Video datasets with consent and data-use approvals
- Licence review
- Constraints
- Non-commercial research and evaluation licence
- Cost figures are projections at an assumed $10 per GPU-hour
- Mixed results on public benchmarks
- Security
- Robot and human-demonstration video can contain identifiable people; apply IRB and data-use rules.
- Proof of value
- Hand-check Flora labels on 50 lab videos
- Measure GPU-hours per video hour locally
Delivery
Role takeaway
- Work
- Run Flora on a lab dataset sample and compare labels and cost with the current method.
- Dependencies
- Research computing
- Licensing office
- IRB where human video is used
- Ownership
- The principal investigator owns use; the licensing office owns terms.
- Skills and adoption
- Graduate researchers need batch-inference and evaluation skills.
- Governance checkpoints
- Licence review before use in sponsored or commercial projects
- Data-use approval for human video
- Acceptance criteria
- Proposed: label quality on a hand-checked sample meets the lab's threshold
- Proposed: measured cost per video hour on local GPUs documented
- Risks
- Licence breach in sponsored work
- Benchmark results not transferring to lab data
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
Labs can run open 9B and 27B models as batch annotation jobs on their own GPUs instead of paying per-call API fees.
Governance
Who approves, reviews and stays accountable for outcomes?
Licence terms (non-commercial research and evaluation) constrain use in sponsored or commercial work.
Security and privacy
What data, permissions and controls need testing?
Human demonstration video may include identifiable people, which brings IRB and data-use obligations.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
Could reduce manual labelling work done by student annotators.
Procurement
What should contracts, pricing and exit terms secure?
No purchase needed for weights, but GPU capacity is the real cost; compare with commercial labelling services on the lab's own data.
Operating model
Which teams own the service once it runs?
Shifts labelling from paid services to locally run batch jobs managed by research computing.
Publication history
- 2026-10-25NVIDIA · Issue 508 resources
Stable resource ID: nvidia-research-flora-open-robotics-data-engine-20261025