# New Release: Ultralytics v8.4.189

**URL:** <https://community.ultralytics.com/t/new-release-ultralytics-v8-4-189/2284>\
**Category:** Discussion\
**Tags:** releases, announcements, ultralytics-official\
**Created:** [October 26, 2026, 2:43pm UTC](https://community.ultralytics.com/t/new-release-ultralytics-v8-4-189/2284 "2026-10-26T14:43:17Z")\
**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:** [October 26, 2026, 2:43pm UTC](https://community.ultralytics.com/t/new-release-ultralytics-v8-4-189/2284/1 "2026-10-26T14:43:17Z")

</div>

# 🚀 Ultralytics v8.4.189: More Reliable Exports and Robust Inference

**Ultralytics YOLO v8.4.189 is here** , bringing deployment fixes for Sony IMX500 and Hailo, broader ONNX export compatibility, clearer dataset errors, and safer inference results handling. **No model architecture changes were introduced** in this release.

## 📦 Export and Deployment Improvements

- **More reliable Sony IMX500 exports:** Updated converter requirements and aligned dependency versions help reduce setup and export failures. IMX500 export is now explicitly supported on **Linux with Python 3.10–3.12**. See [PR #26883](https://github.com/ultralytics/ultralytics/pull/26883) by [@glenn-jocher](https://github.com/ultralytics/ultralytics/pull/26883) for details.
- **Fixed Hailo segmentation exports:** The segmentation prototype output node is now detected from the exported model, accommodating different model activations. This fix is also included in [PR #26883](https://github.com/ultralytics/ultralytics/pull/26883) by [@glenn-jocher](https://github.com/ultralytics/ultralytics/pull/26883).
- **Broader ONNX compatibility:** Replacing several `einsum` operations with equivalent matrix multiplications enables affected YOLOE and YOLO-World models to export at lower ONNX opsets. See [PR #26881](https://github.com/ultralytics/ultralytics/pull/26881) by [@Y-T-G](https://github.com/ultralytics/ultralytics/pull/26881) for the changes.
- **Safer TensorFlow-family exports:** Existing ONNX Runtime variants, including Windows ML and Vitis AI builds, are recognized to avoid unnecessary runtime replacement. See [PR #26870](https://github.com/ultralytics/ultralytics/pull/26870) by [@cainiao33](https://github.com/ultralytics/ultralytics/pull/26870) for details.

## 🛠 Inference and Dataset Fixes

- **Stable classification results:** Saved probabilities remain unchanged across inference calls, even when a backend reuses its output buffer. See [PR #26860](https://github.com/ultralytics/ultralytics/pull/26860) by [@aswanth-07](https://github.com/ultralytics/ultralytics/pull/26860) for the fix.
- **Safer grayscale handling:** Two-dimensional grayscale NumPy images no longer trigger an `IndexError` in `Annotator` and `save_one_box`. See [PR #26878](https://github.com/ultralytics/ultralytics/pull/26878) by [@Rahulbiradar9](https://github.com/ultralytics/ultralytics/pull/26878) for details.
- **More resilient results utilities:** `Results.save_crop` and `verbose` handle results without boxes more safely. See [PR #26868](https://github.com/ultralytics/ultralytics/pull/26868) by [@wizzseen](https://github.com/ultralytics/ultralytics/pull/26868) for the changes.
- **Clearer dataset troubleshooting:** Unreadable datasets are less likely to be misleadingly reported as having no labels, thanks to [PR #26883](https://github.com/ultralytics/ultralytics/pull/26883) by [@glenn-jocher](https://github.com/ultralytics/ultralytics/pull/26883). Classification dataset URL scheme handling is also fixed in [PR #26879](https://github.com/ultralytics/ultralytics/pull/26879) by [@Vaishnavi220506](https://github.com/ultralytics/ultralytics/pull/26879).

## 📱 Documentation and Benchmarks

- **Mobile app guidance:** Ultralytics Platform documentation now explains how to open compatible exports in the Ultralytics YOLO app. See [PR #26875](https://github.com/ultralytics/ultralytics/pull/26875) by [@glenn-jocher](https://github.com/ultralytics/ultralytics/pull/26875) for the update.
- **Refreshed mobile benchmarks:** LiteRT, QNN, and Core ML performance tables now use Ultralytics 8.4.188 assets. See [PR #26874](https://github.com/ultralytics/ultralytics/pull/26874) by [@glenn-jocher](https://github.com/ultralytics/ultralytics/pull/26874) for the measurements.
- **Clearer optimizer guidance:** Documentation explains momentum behavior when using `optimizer=auto`. See [PR #26877](https://github.com/ultralytics/ultralytics/pull/26877) by [@mh-masum6875](https://github.com/ultralytics/ultralytics/pull/26877)—and welcome to this first-time contributor! 🎉

## ⚡ Try It and Share Your Feedback

Install this release with:

```bash
pip install --upgrade ultralytics==8.4.189

```

Explore the [v8.4.189 release notes](https://github.com/ultralytics/ultralytics/releases/tag/v8.4.189) for the overview, or review the [full changelog](https://github.com/ultralytics/ultralytics/compare/v8.4.188...v8.4.189) to see every change.

**How does v8.4.189 work with your models and deployment targets?** Share your results, questions, or issues below—your feedback helps improve future releases. 🙌
