OpenShot 4.0 Ships With a Native Qt6 Timeline, Local AI Object Masking, and Professional Color Grading Tools
OpenShot 4.0 adds screen recording, offline AI subject masking, and a color-grading workspace, completing a Qt5-to-Qt6 rewrite.
Overview
The open-source video editor OpenShot reached version 4.0 on August 30, 2026, an update that OpenShot’s release announcement describes as bringing “some of the biggest creative workflow upgrades” in the project’s history. The release adds built-in screen and webcam recording, a dedicated color-grading workspace, ten new video effects, locally run AI-powered subject masking, and a fully native timeline, according to OpenShot’s announcement and confirmed by LinuxIac. The update marks the first major release since version 3.5.1, according to NotebookCheck, and OpenShot 4.0 was tagged on GitHub the same day, according to Phoronix.
What We Know
A Qt6 rewrite underpins the release
As Phoronix put it: “The main change with OpenShot 4.0 and warranting the bump from OpenShot 3.5.1 is that the Qt6 toolkit is now being utilized.” OpenShot has been ported from Qt5 to Qt6. OpenShot’s own release notes frame this as a broader modernization effort, saying expanded Qt 6 support improves compatibility with newer Linux distributions and lays groundwork for possible future platforms, according to OpenShot. Alongside the interface toolkit, the underlying media libraries also reached new milestones: libopenshot and libopenshot-audio both reached version 1.0.0, according to OpenShot. The update is a free, GPLv3-licensed application available for Windows, macOS, and Linux, according to NotebookCheck.
Built-in recording and a native timeline
OpenShot 4.0 introduces a new Recording View that lets users capture microphone, screen, webcam, and system audio directly into a project, with each source kept as its own separate, editable media file and timeline clip, according to OpenShot. On Linux, recording supports both X11 and Wayland, with PipeWire handling capture on modern Wayland desktops, according to LinuxIac.
The editor’s timeline itself has also changed. According to OpenShot, “OpenShot 4.0 completes the move away from the old web-based timeline components,” with clips, transitions, keyframes, tracks, rulers, and menus now handled through a native Qt timeline, and the current timecode in the ruler can now be edited directly. In OpenShot’s own benchmark comparisons against version 3.5.1, the Blur effect completed its test workload 61.8% faster, Sharpen improved by 12.8%, and timeline rendering improved by 3.4%, rising to 5.1% in a timeline test involving transforms, according to OpenShot.
Color grading and ten new effects
The release adds a new Color View combining a video preview, clip properties, color wheels, and live video scopes, according to OpenShot. Its new Color Grade effect includes controls for exposure, contrast, temperature, tint, highlights, shadows, saturation, and vibrance, plus separate color wheels for global, shadow, midtone, and highlight adjustments and four editable curves for the full image and individual red, green, and blue channels, according to OpenShot. It also supports industry-standard .cube LUT files with adjustable blend intensity, and the entire effect is keyframable, according to OpenShot. New Luma Waveform, Histogram, RGB Parade, and Vectorscope tools round out the color workspace, with the Vectorscope including a skin-tone reference line, according to OpenShot.
OpenShot 4.0 adds ten new effects in total: Audio Visualization, Beat Sync, Film Grain, Denoise Image, Shadow, Glow, Displacement Map, Timer, Color Grade, and Object Mask, according to LinuxIac. The Audio Visualization effect can render waveforms, bars, radial patterns, spectrum displays, particles, and VU-style meters, and in OpenShot’s own performance testing every visualization mode ran above common 24, 30, and 60 FPS project rates, with the most demanding modes exceeding 170 frames per second, according to OpenShot.
Local, offline AI subject masking
The new Object Mask effect lets users mark a subject with a few points and have OpenShot generate and track a selection mask across a clip, and this processing “runs locally,” according to LinuxIac. OpenShot’s release notes specify that the feature uses YOLO, EfficientSAM, and Cutie machine-learning models converted to the ONNX format, adding: “There is no cloud processing requirement, no AI account, and no AI subscription,” according to OpenShot. Object Detection was also updated to support downloadable YOLOv5 ONNX models, model validation, and segmentation masks, according to OpenShot and confirmed by LinuxIac.
What We Don’t Know
OpenShot’s release notes say the 4.0 codebase now includes “important Android compatibility foundations for file access, rendering, Qt bindings, and large ARM64 memory addresses,” but explicitly state that Android is not being announced as a supported release platform and that no release date has been set, according to OpenShot. It remains unclear when, or whether, a supported Android build will follow.