Models: <a href=\"https://huggingface.co/Ultralytics/YOLO26\">https://huggingface.co/Ultralytics/YOLO26</a></p>\n","updatedAt":"2026-06-03T16:34:27.041Z","author":{"_id":"5f1158120c833276f61f1a84","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1608042047613-5f1158120c833276f61f1a84.jpeg","fullname":"Niels Rogge","name":"nielsr","type":"user","isPro":false,"isHf":true,"isHfAdmin":false,"isMod":false,"followerCount":1222,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.44436338543891907},"editors":["nielsr"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/1608042047613-5f1158120c833276f61f1a84.jpeg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2606.03748","authors":[{"_id":"6a1f9080e292c1c78ecb132f","name":"Glenn Jocher","hidden":false},{"_id":"6a1f9080e292c1c78ecb1330","name":"Jing Qiu","hidden":false},{"_id":"6a1f9080e292c1c78ecb1331","name":"Mengyu Liu","hidden":false},{"_id":"6a1f9080e292c1c78ecb1332","name":"Shuai Lyu","hidden":false},{"_id":"6a1f9080e292c1c78ecb1333","name":"Fatih Cagatay Akyon","hidden":false},{"_id":"6a1f9080e292c1c78ecb1334","name":"Muhammet Esat Kalfaoglu","hidden":false}],"publishedAt":"2026-06-02T00:00:00.000Z","submittedOnDailyAt":"2026-06-03T00:00:00.000Z","title":"Ultralytics YOLO26: Unified Real-Time End-to-End Vision Models","submittedOnDailyBy":{"_id":"5f1158120c833276f61f1a84","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1608042047613-5f1158120c833276f61f1a84.jpeg","isPro":false,"fullname":"Niels Rogge","user":"nielsr","type":"user","name":"nielsr"},"summary":"Real-time vision demands models that are accurate, efficient, and simple to deploy across diverse hardware. The YOLO family has become widely deployed for this reason, yet most YOLO detectors still rely on non-maximum suppression at inference, carry heavy detection heads due to Distribution Focal Loss, require long training schedules, and can leave the smallest objects without positive label assignments. We present Ultralytics YOLO26, a unified real-time vision model family that addresses these limitations through coordinated architecture and training advances. YOLO26 uses a dual-head design for native NMS-free end-to-end inference and removes DFL entirely, yielding a lighter head with unconstrained regression range. Its training pipeline combines MuSGD, a hybrid Muon-SGD optimizer adapted from large language model training; Progressive Loss, which shifts supervision toward the inference-time head; and STAL, a label assignment strategy that guarantees positive coverage for small objects. Beyond detection, YOLO26 introduces task-specific head and loss designs for instance segmentation, pose estimation, and oriented detection, producing consistent gains across tasks and scales. The family spans five scales (n/s/m/l/x) and supports detection, instance segmentation, pose estimation, classification, and oriented detection in a single pipeline, with an open-vocabulary extension, YOLOE-26, for text-, visual-, and prompt-free inference. Across all scales, YOLO26 achieves 40.9-57.5 mAP on COCO at 1.7-11.8 ms T4 TensorRT latency, advancing the accuracy-latency Pareto front over prior real-time detectors, while YOLOE-26x reaches 40.6 AP on LVIS minival under text prompting. Code and models are available at https://github.com/ultralytics/ultralytics.","upvotes":2,"discussionId":"6a1f9080e292c1c78ecb1335","projectPage":"https://docs.ultralytics.com/models/yolo26","githubRepo":"https://github.com/ultralytics/ultralytics","githubRepoAddedBy":"user","ai_summary":"YOLO26 addresses real-time vision challenges through a unified model family with NMS-free inference, improved training strategies, and multi-task capabilities spanning detection, segmentation, and pose estimation.","ai_keywords":["YOLO","non-maximum suppression","Distribution Focal Loss","MuSGD","hybrid Muon-SGD optimizer","Progressive Loss","STAL","instance segmentation","pose estimation","oriented detection","open-vocabulary extension","TensorRT latency","mAP","COCO","LVIS"],"ai_summary_model":"Qwen/Qwen2.5-Coder-32B-Instruct","githubStars":57951,"organization":{"_id":"65ba54e4b6b3191a83624cee","name":"Ultralytics","fullname":"Ultralytics","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/60ad037a306d6873ec42d537/iusy6Ia8JeCK-ui4z-kjE.png"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"69830311893522640efeabb0","avatarUrl":"/avatars/951f021d2f8dec9f7e2cf56f737bb069.svg","isPro":false,"fullname":"Jonathan H. Parker","user":"reply-guy","type":"user"},{"_id":"69ccb6d50e56766c20f5e835","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/noauth/syOk9GLF3FCly5Ra3M959.png","isPro":false,"fullname":"Борисов Татьяна","user":"abigaillee8","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"organization":{"_id":"65ba54e4b6b3191a83624cee","name":"Ultralytics","fullname":"Ultralytics","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/60ad037a306d6873ec42d537/iusy6Ia8JeCK-ui4z-kjE.png"},"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2606/2606.03748.md"}">
Ultralytics YOLO26: Unified Real-Time End-to-End Vision Models
Abstract
YOLO26 addresses real-time vision challenges through a unified model family with NMS-free inference, improved training strategies, and multi-task capabilities spanning detection, segmentation, and pose estimation.
Real-time vision demands models that are accurate, efficient, and simple to deploy across diverse hardware. The YOLO family has become widely deployed for this reason, yet most YOLO detectors still rely on non-maximum suppression at inference, carry heavy detection heads due to Distribution Focal Loss, require long training schedules, and can leave the smallest objects without positive label assignments. We present Ultralytics YOLO26, a unified real-time vision model family that addresses these limitations through coordinated architecture and training advances. YOLO26 uses a dual-head design for native NMS-free end-to-end inference and removes DFL entirely, yielding a lighter head with unconstrained regression range. Its training pipeline combines MuSGD, a hybrid Muon-SGD optimizer adapted from large language model training; Progressive Loss, which shifts supervision toward the inference-time head; and STAL, a label assignment strategy that guarantees positive coverage for small objects. Beyond detection, YOLO26 introduces task-specific head and loss designs for instance segmentation, pose estimation, and oriented detection, producing consistent gains across tasks and scales. The family spans five scales (n/s/m/l/x) and supports detection, instance segmentation, pose estimation, classification, and oriented detection in a single pipeline, with an open-vocabulary extension, YOLOE-26, for text-, visual-, and prompt-free inference. Across all scales, YOLO26 achieves 40.9-57.5 mAP on COCO at 1.7-11.8 ms T4 TensorRT latency, advancing the accuracy-latency Pareto front over prior real-time detectors, while YOLOE-26x reaches 40.6 AP on LVIS minival under text prompting. Code and models are available at https://github.com/ultralytics/ultralytics.
Community
Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.
Tap or paste here to upload images
Cite arxiv.org/abs/2606.03748 in a model README.md to link it from this page.
Cite arxiv.org/abs/2606.03748 in a dataset README.md to link it from this page.
Discussion (0)
Sign in to join the discussion. Free account, 30 seconds — email code or GitHub.
Sign in →No comments yet. Sign in and be the first to say something.