{"name":"WeVisOne-VL","repo":"Tencent/WeVisOne-VL","url":"https://midorreal.com/project/tencent-wevisone-vl","repository":"https://github.com/Tencent/WeVisOne-VL","description":"WeVisOne-VL is a powerful VLM by the WeChat Vision Team at Tencent, unifying object-level and pixel-level perception from points and boxes to masks and alpha mattes.","facts":{"language":{"value":"Jupyter Notebook","source":"https://github.com/Tencent/WeVisOne-VL","checked":"22 Nov 2026"},"releases":{"value":"0","source":"https://github.com/Tencent/WeVisOne-VL/releases","checked":"19 Nov 2026"},"contributors":{"value":"2","source":"https://github.com/Tencent/WeVisOne-VL/graphs/contributors","checked":"19 Nov 2026"},"archived":{"value":"no","source":"https://github.com/Tencent/WeVisOne-VL","checked":"22 Nov 2026"}},"stars":1252.0,"writeup":{"what_it_is":"WeVisOne-VL is a vision-language model from the WeChat Vision Team at Tencent. It handles object-level tasks such as detection, referring localization, counting, pointing and OCR, and pixel-level tasks such as referring segmentation and alpha matting, using a 9.6M-parameter pixel decoder.","audience":"Researchers and developers working on visual perception with vision-language models","claims":[{"kind":"specific","claim":"Unifies detection, referring, counting, pointing, OCR, segmentation and matting in one model","excerpt":"detection, referring, counting, pointing, OCR, segmentation and matting in a single model","status":"not_checked"},{"kind":"specific","claim":"Uses a 9.6M-parameter pixel decoder to output masks and alpha mattes without SAM","excerpt":"a 9.6M-parameter decoder outputs masks and alpha mattes, with no SAM required","status":"not_checked"},{"kind":"specific","claim":"Returns None for absent targets instead of hallucinating","excerpt":"returns `None` for absent targets instead of hallucinating","status":"not_checked"},{"kind":"specific","claim":"Achieves 55.3 F1 Mean on LVIS detection for the 4B model","excerpt":"| 📦 Detection       | LVIS                 | F1 Mean ↑   | 50.7          | **55.3**","status":"not_checked"}],"alternatives":[],"written_by":"AI, from the project's README","date":"2026-10-26"},"why_now":null,"owner_supplied":null,"score":{"verdict":"too_early","hype":61,"reality":12,"gap":50,"momentum":null,"confidence":60,"version":"score-2.0","rule":"verdict-2.0","calculated_at":"2026-11-22T07:00:00+00:00","verdict_source":"rule"},"early_read":{"estimate":true,"lean":"too_close","gap_low":10,"gap_high":50,"rule":"early-2.0"},"changes":[{"date":"2026-10-22T08:58:16+00:00","text":"Repository created","source":"https://github.com/Tencent/WeVisOne-VL"}],"cite":"WeVisOne-VL is a GitHub project at https://github.com/Tencent/WeVisOne-VL written mainly in Jupyter Notebook, described there as \"WeVisOne-VL is a powerful VLM by the WeChat Vision Team at Tencent, unifying object-level and pixel-level perception from points and boxes to masks and alpha mattes\".","attribution":"Data from Mid or Real, https://midorreal.com/project/tencent-wevisone-vl, checked 22 November 2026. Attribution with a link is required.","still_real":null}