{"resourceId":"nvidia-research-flora-open-robotics-data-engine-20261025","versions":[{"version":"external-37f249f6f32382192fab1a1c8640af6abc2507e2aef0b02f4eeaee6596d7ff6e","resource":{"id":"nvidia-research-flora-open-robotics-data-engine-20261025","title":"NVIDIA GEAR Lab releases Flora, an open model suite for labelling robot and human video, claiming ~125x lower cost than GPT-6 Astra","organization":"NVIDIA Research (GEAR Lab)","sector":"Research enterprise (robotics and embodied AI)","geography":"Global (open models on Hugging Face)","publishedAt":"Undated project page (paper 2026); covered by The Decoder on 2026-10-25","publicationDate":null,"eventDate":null,"sourceName":"NVIDIA Research","sourceLabel":"Vendor research project page","sourceUrl":"https://research.nvidia.com/labs/gear/flora/","evidenceClass":"vendor-claim","outcomeClass":"mixed","topics":["developers-agents","infrastructure"],"finding":"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).","sledRelevance":"Limited SLED relevance: matters mainly to university robotics and AI research groups evaluating open annotation tools. No public-sector use is described.","evidence":"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.","architectureImplications":"Labs can run open 9B and 27B models as batch annotation jobs on their own GPUs instead of paying per-call API fees.","governanceImplications":"Licence terms (non-commercial research and evaluation) constrain use in sponsored or commercial work.","securityPrivacyImplications":"Human demonstration video may include identifiable people, which brings IRB and data-use obligations.","caveats":"Vendor benchmarks; cost estimates not production measurements; page undated.","streamIds":["nvidia"],"roles":{"sales":"Interpretation: relevant only for university robotics and embodied-AI groups, where open labelling models are a credible reason to discuss research GPU capacity. Present the 125x cost figure as NVIDIA's estimate, note the mixed benchmark table, and flag the non-commercial licence early, because many lab projects have industry sponsors. Treat it as a niche research conversation, not a government opportunity.","engineering":"Interpretation: Flora is three fine-tuned Qwen3.5 models (9B segmenter, two 27B models) suited to batch jobs on research GPU clusters. Pre-sales should size GPU-hours per video hour on the customer's hardware rather than rely on NVIDIA's $10 per GPU-hour assumption, and benchmark on the lab's own robot data, where results trailed GPT-6 Astra. The 27B models need multi-GPU capacity or quantisation, which affects which campus clusters can run them.","delivery":"Interpretation: delivery is a short evaluation: run Flora on a hand-checked sample, measure label quality and actual GPU cost, and complete a licence review before any sponsored use. Proposed acceptance criteria are a documented quality threshold and measured cost per video hour. Check data-use approvals for any human demonstration video before running the pipeline, since recordings can include identifiable people."},"roleDetails":{"sales":{"customerProblem":"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"],"discoveryQuestions":["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?"],"valueHypothesis":"Interpretation: open annotation models may lower labelling costs for academic robotics labs; savings are unmeasured outside NVIDIA's estimates.","potentialEngagement":"Licensing review and small benchmark of Flora against the lab's current labelling process.","unsupportedClaims":["That labs will see 125x savings","That Flora beats GPT-6 Astra on robot video","That commercial use is allowed"]},"engineering":{"fit":"Academic robotics and embodied-AI research groups with GPU access; weak fit for most state and local government.","architectureIntegration":"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.","proofOfValue":["Hand-check Flora labels on 50 lab videos","Measure GPU-hours per video hour locally"]},"delivery":{"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.","skillsAdoption":"Graduate researchers need batch-inference and evaluation skills.","governanceCheckpoints":["Licence review before use in sponsored or commercial projects","Data-use approval for human video"],"acceptanceCriteria":["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"]}},"retrievedAt":"2026-10-26T03:02:15Z","enrichedAt":"2026-10-26T03:14:13Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Could reduce manual labelling work done by student annotators.","procurementImplications":"No purchase needed for weights, but GPU capacity is the real cost; compare with commercial labelling services on the lab's own data.","operatingModelImplications":"Shifts labelling from paid services to locally run batch jobs managed by research computing.","sourceVerification":{"openedUrl":"https://research.nvidia.com/labs/gear/flora/","referenceExcerpt":"Flora is battle tested at scale, labeling 200,000 hours of data per day across 512 GB300s.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}