{"version":1,"type":"story","url":"https://digestai.news/story/voxel51-launches-3-0-platform-for-physical-ai-data-workflows","json":"https://digestai.news/story/voxel51-launches-3-0-platform-for-physical-ai-data-workflows.json","markdown":"https://digestai.news/story/voxel51-launches-3-0-platform-for-physical-ai-data-workflows.md","slug":"voxel51-launches-3-0-platform-for-physical-ai-data-workflows","headline":"Voxel51 launches 3.0 platform for physical AI data workflows","summary":"Voxel51 introduced **Voxel51 3.0** on October 15, 2026, an end-to-end multimodal data platform designed for teams working on physical AI—including robotics, autonomous systems, manufacturing, and healthcare. The update consolidates search, curation, annotation, and evaluation into a single system, with free self-service access for users to test it on their own data at [app.voxel51.com](https://app.voxel51.com). The company also rebranded its enterprise product from **FiftyOne Enterprise** to **Voxel51**, while keeping **FiftyOne** as the open-source project name.\n\nKey features include **multimodal search** with hybrid retrieval (semantic + lexical) and time-series support for sensor data, **Voxel51 Agent**—a workflow automation tool built on the **Model Context Protocol (MCP)**—and **agentic labeling** for 2D, 3D, video, and multimodal data. The platform claims cost savings of up to **15x** when benchmarking agent tasks across models like **Claude Sonnet 5**, **GPT-5.1**, and **Gemini 3.1 Pro**. Data provenance tracking spans four levels, ensuring auditability and reproducibility. Voxel51 also announced **Nebius Workshops** in Canada and the US through November 2026, offering hands-on sessions on deploying workflows like embeddings and auto-labeling.\n\nThe company cites its **2026 State of Physical AI survey**, which found that **58% of exceptional teams** spend over half their project time on data work, compared to **21% of struggling teams**. Voxel51 positions 3.0 as a model-agnostic solution, allowing teams to integrate their own foundation models and domain-specific tools without vendor lock-in.","keyPoints":["Voxel51 3.0 unifies search, curation, annotation, and evaluation for physical AI teams, with free self-service access","Agentic workflows use 45+ MCP tools, 80+ operators, and 18 open-source skills, with cost savings up to 15x across models","Data provenance spans deployment, dataset, sample, and versioning levels, with Nebius Workshops launching in November 2026"],"whyItMatters":"For robotics and autonomous systems teams, this platform reduces manual data work and integrates multimodal data pipelines, cutting costs and improving reproducibility in physical AI deployment.","category":{"slug":"agents","name":"Agents & Tools","url":"https://digestai.news/category/agents"},"entities":{"companies":["Voxel51","FiftyOne","Nebius","Anthropic","OpenAI","Google"],"models":["Claude Sonnet 5","Claude Opus 5","GPT-5.1","Gemini 3.1 Pro"],"people":["Brian Moore","Ritchie Martori"]},"firstPublishedAt":"2026-10-15T16:22:16Z","updatedAt":"2026-10-15T16:22:16Z","sourceCount":1,"hasPrimarySource":false,"sources":[{"outlet":"Unite.AI","title":"Voxel51 Unveils Unified Data Platform 3.0 Aimed at Physical AI Teams","url":"https://unite.ai/voxel51-unveils-unified-data-platform-3-0-aimed-at-physical-ai-teams","publishedAt":"2026-10-15T16:22:16Z","type":"press","primary":false,"lead":true}],"sourceNotes":null,"discussions":[],"thread":null,"cite":{"text":"Digest AI, \"Voxel51 launches 3.0 platform for physical AI data workflows\", 15 October 2026, https://digestai.news/story/voxel51-launches-3-0-platform-for-physical-ai-data-workflows","publisher":"Digest AI","title":"Voxel51 launches 3.0 platform for physical AI data workflows","datePublished":"2026-10-15T16:22:16Z","url":"https://digestai.news/story/voxel51-launches-3-0-platform-for-physical-ai-data-workflows"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}