{"apiVersion":"v1","taxonomyVersion":1,"generatedAt":"2026-10-29T15:05:43.776Z","data":{"recordId":"wsa_z83b4g9mc9yy","score":0.72,"confidence":"medium","capped":false,"capReason":null,"sources":[{"n":1,"url":"https://huggingface.co/lightonai/LightOnOCR-3-4B","publisher":"Hugging Face","authority":"owner-controlled","verification":"verified","lastRetrievedAt":"2026-10-08T21:36:35.267Z","detectedPublishedAt":"2026-10-06T16:12:40.000Z","rights":"original"},{"n":2,"url":"https://huggingface.co/lightonai/LightOnOCR-3-1B","publisher":"Hugging Face","authority":"owner-controlled","verification":"verified","lastRetrievedAt":"2026-10-29T00:08:40.326Z","detectedPublishedAt":"2026-10-07T09:25:46.000Z","rights":"original"},{"n":3,"url":"https://huggingface.co/lightonai/LightOnOCR-3-0.8B","publisher":"Hugging Face","authority":"owner-controlled","verification":"verified","lastRetrievedAt":"2026-10-29T00:08:40.326Z","detectedPublishedAt":"2026-10-07T10:23:52.000Z","rights":"original"}],"claims":[{"id":"c5d461f6-bb90-4b01-b4e9-f48ac2e4aed7","kind":"capability","statement":"In transcription mode, the model outputs page text as markdown.","citations":[{"source":3,"supportProbability":0.92,"supports":true,"quote":{"text":"* 📝 **Transcription mode:** call the model with an empty prompt and get the full page text, as before.","rights":"original"}}]},{"id":"4cf0b2d1-3726-47c4-8dac-801eb6cf5c89","kind":"capability","statement":"In grounding mode, the model returns content blocks with labels and bounding boxes.","citations":[{"source":3,"supportProbability":0.96,"supports":true,"quote":{"text":"* 📍 **Grounding mode:** call it with the `grounding` prompt and every block comes back with a label and a bounding box","rights":"original"}}]},{"id":"a19dc690-4509-4c1b-854c-e30ab9de5126","kind":"capability","statement":"Image blocks include short descriptions, and chart blocks contain tables of extracted data points.","citations":[{"source":3,"supportProbability":0.95,"supports":true,"quote":{"text":"* **Chart blocks** contain an HTML table of the data points extracted from the figure, turning visual information into structured data.","rights":"original"}}]},{"id":"d5643d9c-86f7-431f-b079-6a5fb269c170","kind":"deployment","statement":"The model can be loaded with Transformers or served with vLLM.","citations":[{"source":3,"supportProbability":0.89,"supports":true,"quote":null}]},{"id":"2d8ea666-5fe5-4e70-b299-da2e739642dd","kind":"capability","statement":"The 0.8B variant is described as fast and efficient.","citations":[{"source":2,"supportProbability":0.98,"supports":true,"quote":{"text":"The 1B keeps the LightOnOCR-2-1B architecture, while the 0.8B and 4B adopt the Qwen3.5 vision-language architecture, which simplifies integration with existing tools and brings a significant speed-up.","rights":"original"}}]},{"id":"db3ad9e8-114d-42ed-991e-12edd8a6d739","kind":"capability","statement":"LightOnOCR-3 models provide bounding boxes and labels for visual document elements.","citations":[{"source":2,"supportProbability":0.57,"supports":false,"quote":{"text":"Compared to the previous generation, they bring significant improvements in speed and transcription quality and introduce new visual understanding features: the models now output bounding box coordinates with labels for all visual elements of a document, short descriptions of…","rights":"original"}}]},{"id":"82113590-57ab-4b7f-804b-86703989ffd7","kind":"capability","statement":"LightOnOCR-3 models can provide image descriptions and extract chart data into tables.","citations":[{"source":2,"supportProbability":0.59,"supports":false,"quote":null}]},{"id":"65482fc1-f4f1-41c9-8079-1132bf6c267c","kind":"availability","statement":"The LightOnOCR-3-0.8B model is linked from the model card.","citations":[{"source":2,"supportProbability":0.53,"supports":false,"quote":null}]},{"id":"bbc02aae-9523-4080-a376-96b35da5a538","kind":"deployment","statement":"The model family includes usage instructions for Transformers and vLLM.","citations":[{"source":2,"supportProbability":0.11,"supports":false,"quote":null}]},{"id":"93fe537a-8fa2-40fb-9bf6-882e961a992b","kind":"pricing","statement":"The model family is released under Apache License 2.0 for research and commercial use.","citations":[{"source":2,"supportProbability":0.84,"supports":true,"quote":{"text":"All models are released under the Apache 2.0 license for research and commercial use.","rights":"original"}}]}]},"nextCursor":null,"license":{"catalog":{"name":"CC BY 4.0","spdx":"CC-BY-4.0","url":"https://creativecommons.org/licenses/by/4.0/","attribution":"wowsuchai","attributionUrl":"https://wowsuchai.withglint.app","covers":["catalog metadata","classifications","generated summaries","feed event metadata"]},"excluded":{"covers":["source quotations","linked source content"],"terms":"Not licensed by wowsuchai. Source quotations and linked source content keep their original rights."}}}