Question about YOLO26n mAP50-95 discrepanc

[Brayan532]

Dear Ultralytics Team,

I hope you’re doing well.

While reviewing the official Ultralytics YOLO26 webpage and comparing it with the Object Detection Model Leaderboard page, I noticed a discrepancy in the reported mAP50-95 for YOLO26n.

On the official YOLO26 page, the value is listed as 40.9 (for the case excluding end-to-end), whereas on the leaderboard page, it appears as 39.9.

Could you please clarify the reason for this difference? It may simply be a typo, but I wanted to check.

Best regards,
Brayan.

[UltralyticsAssistant]

👋 Hello @Brayan532, thank you for your interest in Ultralytics 🚀 and for carefully checking the YOLO26 documentation and leaderboard details. This is an automated response to help get your discussion moving, and an Ultralytics engineer will also assist you soon 😊

We recommend a visit to the Docs for new users where you can find many Python and CLI usage examples and where many of the most common questions may already be answered.

Since this appears to be a documentation or reporting discrepancy question, it would be helpful if you could share the exact page sections, screenshots, or copied table rows you are comparing so the team can review the difference more quickly 🔍

If this is a 🐛 Bug Report, please provide a minimum reproducible example to help us debug it.

If this is a custom training ❓ Question, please provide as much information as possible, including dataset image examples and training logs, and verify you are following our Tips for Best Training Results.

Join the Ultralytics community where it suits you best. For real-time chat, head to Discord 🎧. Prefer in-depth discussions? Check out Discourse. Or dive into threads on our Subreddit to share knowledge with the community.

Upgrade

Upgrade to the latest ultralytics package including all requirements in a Python>=3.8 environment with PyTorch>=1.8 to verify your issue is not already resolved in the latest version:

pip install -U ultralytics

Environments

YOLO may be run in any of the following up-to-date verified environments (with all dependencies including CUDA/CUDNN, Python and PyTorch preinstalled):

• Notebooks with free GPU:
• Google Cloud Deep Learning VM. See GCP Quickstart Guide
• Amazon Deep Learning AMI. See AWS Quickstart Guide
• Docker Image. See Docker Quickstart Guide

Status

If this badge is green, all Ultralytics CI tests are currently passing. CI tests verify correct operation of all YOLO Modes and Tasks on macOS, Windows, and Ubuntu every 24 hours and on every commit.

[glenn-jocher]

Thanks for catching this — the current YOLO26 model page lists YOLO26n at 40.9 mAP50-95, with 40.1 shown separately for e2e, so the 39.9 on the leaderboard looks like a stale entry rather than a different evaluation setting. We’ll get that synced.

[Y-T-G]

@Brayan532 The leaderboard page is not by Ultralytics. You would have to report the discrepancy to the publisher of the page. For correct metrics, refer to the metrics published in the Ultralytics YOLO26 page:

https://docs.ultralytics.com/models/yolo26/#performance-metrics