# New Release: Ultralytics v8.4.107

**URL:** <https://community.ultralytics.com/t/new-release-ultralytics-v8-4-107/2137>\
**Category:** Discussion\
**Tags:** ultralytics-official, releases, announcements\
**Created:** [July 27, 2026, 1:49pm UTC](https://community.ultralytics.com/t/new-release-ultralytics-v8-4-107/2137 "2026-07-27T13:49:27Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![glenn-jocher](https://sea1.discourse-cdn.com/flex001/user_avatar/community.ultralytics.com/glenn-jocher/32/83_2.png) [@glenn-jocher](https://community.ultralytics.com/u/glenn-jocher)\
**Post date:** [July 27, 2026, 1:49pm UTC](https://community.ultralytics.com/t/new-release-ultralytics-v8-4-107/2137/1 "2026-07-27T13:49:27Z")

</div>

# 🚀 Ultralytics v8.4.107: Huawei Ascend Support Is Here

## 🌟 Quick Overview

The [Ultralytics v8.4.107 release](https://github.com/ultralytics/ultralytics/releases/tag/v8.4.107) introduces hardware-optimized Huawei Ascend export and inference, faster FLOPs profiling, more reliable distributed training, and important tracking, evaluation, and dataset fixes.

The headline feature: Ultralytics YOLO models can now be compiled into `.om` models and accelerated on Huawei Ascend NPUs. 🧠⚡

## 🆕 New Features

### Huawei Ascend Export and Inference

In [PR #25432](https://github.com/ultralytics/ultralytics/pull/25432), [@glenn-jocher](https://github.com/glenn-jocher) added `format=ascend` with:

- Export through Huawei’s CANN ATC compiler.
- Self-describing `.om` models containing deployment metadata.
- Target selection through `name`, including `Ascend310P3` and `Ascend310B4`.
- `AutoBackend` inference powered by `ais_bench`.
- Support for detection, segmentation, classification, pose, OBB, semantic segmentation, and depth models.
- Static-shape FP16 compilation.
- Host-side export without requiring an attached Ascend device.
- Format aliases including `huawei`, `cann`, and `om`.
- Correct routing so `cann` is no longer interpreted as NCNN.

This creates a direct deployment path for Huawei Atlas boards, OrangePi AIPro devices, industrial inspection systems, robotics, and smart cameras. 🏭🤖

### Ultralytics Platform Integration ☁

Huawei Ascend increases the available deployment formats from 19 to 20 across the [Ultralytics Platform](https://platform.ultralytics.com), making the new export option available throughout format-selection and deployment workflows.

## ⚡ Improvements

### Faster, Cleaner FLOPs Profiling

In [PR #25434](https://github.com/ultralytics/ultralytics/pull/25434), [@glenn-jocher](https://github.com/glenn-jocher) changed FLOPs profiling to operate directly on models rather than creating expensive deep copies, reducing time and memory usage.

In [PR #25438](https://github.com/ultralytics/ultralytics/pull/25438), [@glenn-jocher](https://github.com/glenn-jocher) removed the unused `get_flops_with_torch_profiler` utility and updated the requirement to `ultralytics-thop>=2.1.0` for improved performance and cleanup reliability.

### More Reliable Distributed Training 🌐

In [PR #25435](https://github.com/ultralytics/ultralytics/pull/25435), [@glenn-jocher](https://github.com/glenn-jocher) updated DDP rendezvous port selection to prefer random available ports below the operating system’s ephemeral range, reducing failures caused by port reuse while retaining an OS fallback.

### Clearer Disk-Space Reporting 💾

In [PR #25433](https://github.com/ultralytics/ultralytics/pull/25433), [@glenn-jocher](https://github.com/glenn-jocher) improved low-disk-space diagnostics by reporting available capacity in KB when less than 1 MB remains, instead of showing an ambiguous `0.0 MB`.

## 🛠 Bug Fixes

- In [PR #25418](https://github.com/ultralytics/ultralytics/pull/25418), [@JESUSROYETH](https://github.com/JESUSROYETH) restored TrackTrack deleted-detection recovery for PyTorch models.
- In [PR #25419](https://github.com/ultralytics/ultralytics/pull/25419), [@JESUSROYETH](https://github.com/JESUSROYETH) corrected class-label and AP alignment in `PR_curve.png` when a class has no predictions.
- In [PR #25420](https://github.com/ultralytics/ultralytics/pull/25420), [@JESUSROYETH](https://github.com/JESUSROYETH) prevented empty segmentation annotations from corrupting bounding boxes in `GroundingDataset`.
- In [PR #25417](https://github.com/ultralytics/ultralytics/pull/25417), [@JESUSROYETH](https://github.com/JESUSROYETH) made optimizer-name handling case-insensitive before MuSGD and automatic optimizer checks, so values such as `musgd` now work correctly.

## 🚀 Try Huawei Ascend Export

Upgrade Ultralytics:

```bash
pip install -U ultralytics

```

Export a YOLO26 model for an Ascend target:

```bash
yolo export model=yolo26n.pt format=ascend name=Ascend310B4

```

The generated model directory can then be loaded through the standard Ultralytics prediction APIs.

> ⚠ Ascend export requires the Linux-based CANN toolkit and the `atc` compiler. Ascend inference also requires the CANN runtime and `ais_bench` on the target device.

## 🙌 Get Started and Share Feedback

Please try v8.4.107 with your training, export, tracking, and Ascend deployment workflows. Feedback, reproducible issue reports, and deployment results are always appreciated by the Ultralytics team and YOLO community.

For every change included in this release, review the [complete v8.4.106 to v8.4.107 changelog](https://github.com/ultralytics/ultralytics/compare/v8.4.106...v8.4.107) on GitHub.