# New Release: Ultralytics v8.4.114 **URL:** \ **Category:** Discussion\ **Tags:** ultralytics-official, releases, announcements\ **Created:** [August 1, 2026, 1:31pm UTC](https://community.ultralytics.com/t/new-release-ultralytics-v8-4-114/2150 "2026-08-01T13:31:34Z")\ **Posts on this page:** 1\ **Page:** 1
**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:** [August 1, 2026, 1:31pm UTC](https://community.ultralytics.com/t/new-release-ultralytics-v8-4-114/2150/1 "2026-08-01T13:31:34Z")
# 🚀 Ultralytics v8.4.114 Released Ultralytics v8.4.114 improves reliability across **Ultralytics Platform workflows, exported-model validation, edge inference, training pipelines, and advanced vision tasks**. Expect clearer errors, simpler validation, faster LiteRT CPU inference, and more dependable deployment workflows. ⚡ ## 🌟 Highlights - **Clearer Platform errors:** Actionable API messages now surface correctly, while invalid credentials and failed console uploads stop unnecessary retries. - **Simpler exported-model validation:** Static exports automatically reuse the image size stored in their metadata. - **Faster edge inference:** LiteRT now uses configured CPU threads for multi-core execution, with updated Raspberry Pi 5 benchmarks for YOLO26n and YOLO26s. - **More reliable exports:** TorchScript, NCNN, and Paddle workflows receive important compatibility fixes. - **Stronger vision pipelines:** Fixes cover SAM3, pose rendering, class activation maps, semantic segmentation, depth training, and distributed validation. ## 🆕 New Features and Performance - [PR #25562](https://github.com/ultralytics/ultralytics/pull/25562) by [@lakshanthad](https://github.com/lakshanthad) enables multi-core Google LiteRT CPU inference using the configured thread count. - [PR #25564](https://github.com/ultralytics/ultralytics/pull/25564) by [@lakshanthad](https://github.com/lakshanthad) updates Raspberry Pi 5 LiteRT benchmarks for YOLO26n and YOLO26s, reflecting substantially lower latency. - [PR #25559](https://github.com/ultralytics/ultralytics/pull/25559) by [@synml](https://github.com/synml) reuses `imgsz` from export metadata when validating static ONNX, TensorRT, OpenVINO, and similar models. ✅ ## 🛠 Platform and Deployment Improvements - [PR #25581](https://github.com/ultralytics/ultralytics/pull/25581) by [@glenn-jocher](https://github.com/glenn-jocher) surfaces detailed Platform errors, stops requests early when no API key is available, disables further attempts after invalid credentials, and prevents console-output upload failures from creating retry loops. - [PR #25568](https://github.com/ultralytics/ultralytics/pull/25568) by [@Y-T-G](https://github.com/Y-T-G) fixes NCNN and Paddle export for prompt-free YOLOE models. - [PR #25569](https://github.com/ultralytics/ultralytics/pull/25569) by [@Y-T-G](https://github.com/Y-T-G) resolves TorchScript GPU inference device mismatches by keeping generated anchors on the runtime device. ## 🐛 Bug Fixes - [PR #25563](https://github.com/ultralytics/ultralytics/pull/25563) by [@JESUSROYETH](https://github.com/JESUSROYETH) defines the required `mask_threshold` for SAM3 semantic prediction, preventing an `AttributeError`. - [PR #25582](https://github.com/ultralytics/ultralytics/pull/25582) by [@leoventuroso](https://github.com/leoventuroso) prevents `class_activation_map` from raising an `IndexError` for out-of-range class IDs. - [PR #25565](https://github.com/ultralytics/ultralytics/pull/25565) by [@JESUSROYETH](https://github.com/JESUSROYETH) ensures `Annotator.kpts()` renders keypoints and limbs located exactly on image borders. - [PR #25567](https://github.com/ultralytics/ultralytics/pull/25567) by [@Zenka737](https://github.com/Zenka737) fixes analytics line-chart counts resetting before `update_graph` consumes them. - [PR #25576](https://github.com/ultralytics/ultralytics/pull/25576) by [@Y-T-G](https://github.com/Y-T-G) prevents distributed validation failures when the total batch size exceeds the number of validation images. - [PR #25580](https://github.com/ultralytics/ultralytics/pull/25580) by [@raimbekovm](https://github.com/raimbekovm) applies BGR augmentation correctly during semantic segmentation and depth training. ## 📚 Documentation and Maintenance - [PR #25561](https://github.com/ultralytics/ultralytics/pull/25561) by [@synml](https://github.com/synml) documents `channels_last` in the validation arguments table. - [PR #25577](https://github.com/ultralytics/ultralytics/pull/25577) by [@raimbekovm](https://github.com/raimbekovm) documents depth-loss gains and updates semantic and depth argument tables. - [PR #25574](https://github.com/ultralytics/ultralytics/pull/25574) by [@onuralpszr](https://github.com/onuralpszr) fixes installation links in the Intel DL Streamer documentation. - [PR #25579](https://github.com/ultralytics/ultralytics/pull/25579) by [@RizwanMunawar](https://github.com/RizwanMunawar) corrects a Python syntax error in the similarity-search solution. - [PR #25578](https://github.com/ultralytics/ultralytics/pull/25578) by [@Laughing-q](https://github.com/Laughing-q) removes redundant regression tests to reduce test-suite maintenance overhead. ## 👋 New Contributor A warm welcome to [@leoventuroso](https://github.com/leoventuroso), who made their first contribution in [PR #25582](https://github.com/ultralytics/ultralytics/pull/25582). Thank you! 🎉 ## 📦 Try It Today Upgrade to the latest release with: ```bash pip install -U ultralytics ``` Visit the [Ultralytics v8.4.114 release page](https://github.com/ultralytics/ultralytics/releases/tag/v8.4.114) for the release details, or review the [complete v8.4.113 to v8.4.114 changelog](https://github.com/ultralytics/ultralytics/compare/v8.4.113...v8.4.114) to inspect every change. Please try the release and share your feedback, benchmarks, and any issues you encounter. Thanks to every contributor and the wider YOLO community for helping make Ultralytics YOLO faster and more reliable! 🙌