{"apiVersion":"v1","taxonomyVersion":1,"generatedAt":"2026-10-29T14:07:16.672Z","data":{"recordId":"wsa_djgt5caauhme","score":0.844,"confidence":"high","capped":false,"capReason":null,"sources":[{"n":1,"url":"https://github.com/Tencent/WeVisOne-VL","publisher":"GitHub new repositories","authority":"owner-controlled","verification":"verified","lastRetrievedAt":"2026-10-29T00:46:41.159Z","detectedPublishedAt":"2026-10-22T08:58:16.000Z","rights":"original"},{"n":2,"url":"https://huggingface.co/tencent/WeVisOne-VL-2B","publisher":"Tencent Hunyuan","authority":"owner-controlled","verification":"verified","lastRetrievedAt":"2026-10-29T00:46:41.159Z","detectedPublishedAt":"2026-10-22T09:47:50.000Z","rights":"original"}],"claims":[{"id":"c213f974-8b86-4790-a463-faffa6f20c57","kind":"capability","statement":"The pixel decoder outputs masks and alpha mattes without requiring SAM.","citations":[{"source":1,"supportProbability":0.51,"supports":false,"quote":{"text":"- 🪶 **Lightweight Pixel Decoder**: a 9.6M-parameter decoder outputs masks and alpha mattes, with no SAM required.","rights":"original"}}]},{"id":"840ba826-f548-44c5-abff-25c3a9d2e4d1","kind":"availability","statement":"Model weights, an inference package, cookbooks, and an evaluation pipeline are released.","citations":[{"source":1,"supportProbability":0.55,"supports":false,"quote":{"text":"- **[2026-10]** 🎉 We release the technical report, model weights (2B & 4B), inference package, Cookbooks, and evaluation pipeline!","rights":"original"}}]},{"id":"5d6fc6b7-b15a-4a11-a015-ab975aa207a5","kind":"availability","statement":"A 2B demo is available on Hugging Face Spaces.","citations":[{"source":1,"supportProbability":0.84,"supports":true,"quote":{"text":"- [x] Model weights (2B & 4B) on Hugging Face","rights":"original"}}]},{"id":"fadcfc73-e8b2-4ba2-9f7c-6fea31100fbb","kind":"capability","statement":"The model supports open-vocabulary object detection, referring localization, counting, pointing, and OCR.","citations":[{"source":2,"supportProbability":0.85,"supports":true,"quote":{"text":"Built upon Qwen3-VL, it unifies object-level perception (open-vocabulary object detection, referring localization, object counting, object pointing, and OCR) and pixel-level perception (referring segmentation and promptable alpha matting) through structured next-token…","rights":"original"}}]},{"id":"6a0146d4-ea1f-4941-8f39-db4296a5bad4","kind":"capability","statement":"The model supports referring segmentation and promptable alpha matting.","citations":[{"source":2,"supportProbability":0.91,"supports":true,"quote":{"text":"Built upon Qwen3-VL, it unifies object-level perception (open-vocabulary object detection, referring localization, object counting, object pointing, and OCR) and pixel-level perception (referring segmentation and promptable alpha matting) through structured next-token…","rights":"original"}}]},{"id":"d0fe823e-19b4-46b4-a07b-39f66d9280e5","kind":"availability","statement":"The Quick Start loads the model from tencent/WeVisOne-VL-2B and uses the wevisone_vl inference package.","citations":[{"source":2,"supportProbability":0.92,"supports":true,"quote":null}]},{"id":"c0275af7-8afa-46ef-a4a2-53dd7643bdaf","kind":"deployment","statement":"The example initializes the model with device=\"auto\" for inference.","citations":[{"source":2,"supportProbability":0.95,"supports":true,"quote":null}]},{"id":"9127441c-95dd-4a62-a04c-edf2fbc8b88d","kind":"recommendation-relevant","statement":"The release is intended for research and is not intended for autonomous high-impact decisions or identifying individuals.","citations":[{"source":2,"supportProbability":0.93,"supports":true,"quote":{"text":"The release is intended for research in visual grounding, referring localization, pointing, counting, OCR localization, referring segmentation, and image matting.","rights":"original"}}]},{"id":"edb32062-d961-4382-8695-2a38f3c21f16","kind":"recommendation-relevant","statement":"The model currently targets single-turn image perception, and video grids are rejected by the pixel decoder.","citations":[{"source":2,"supportProbability":0.94,"supports":true,"quote":{"text":"- For inference efficiency, the model directly generates structured outputs without explicit chain-of-thought (CoT) reasoning, and currently targets single-turn image perception (video grids are rejected by the pixel decoder).","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."}}}