Ai

Deep-Live-Cam 2.7 Beta Adds Realtime Face Enhancement, Multi-GPU Support, and Faster Performance

Deep-Live-Cam 2.7 Beta Adds Realtime Face Enhancement, Multi-GPU Support, and Faster Performance

Deep-Live-Cam 2.7-beta is the project’s biggest update so far, centered on major improvements to realtime video processing, face enhancement, masking controls, and GPU performance. The release introduces a new realtime face enhancer, expanded GPEN model support, multi-GPU switching, LUT-based grading tools, and a range of workflow upgrades that make live preview and window projection more practical for continuous use.

What Changed

The headline addition in version 2.7-beta is a realtime face enhancer, which is paired with an inswapper optimizer and two new face enhancement models, GPEN 512 and GPEN 256. The update also adds a face enhancer scaler, giving users more flexibility in tuning output quality and performance for live workloads.

Masking and facial refinement also received a substantial upgrade. New and improved controls include quick lip mask, lip mask, chin mask, and eyes mask, along with interpolation features designed to improve motion consistency and overall realism in processed video streams.

On the hardware side, the release adds a GPU changer for multi-GPU support, which is especially useful for users running more advanced workstations. The changelog also highlights performance gains, with the project claiming 27 frames per second even while face enhancement is enabled.

Visual workflow improvements are another major part of the release. Version 2.7-beta adds LUT support for easier grading, window projection, in-window preview, realtime video watching, and a fullscreen mode for watching projected realtime videos. Additional usability improvements include camera refresh, a resolution changer, and a broader UI refresh.

Why It Matters

This version matters because it moves Deep-Live-Cam further toward a more complete live production workflow rather than just a point-feature update. The combination of realtime enhancement, better masking tools, and expanded preview options suggests a stronger focus on creators and operators who need immediate visual feedback and more control during live sessions.

The multi-GPU support and reported performance improvements are also notable for users pushing heavier realtime pipelines. Together with LUT grading and fullscreen projection, the update broadens the software’s utility in more advanced desktop and streaming environments.

While the announcement includes promotional language and external website links, the core release notes point to a substantial feature expansion in this beta. For users tracking the project, the most important takeaway is that version 2.7-beta appears to be a major functional leap, especially in realtime enhancement quality, system flexibility, and live viewing workflow.

Official Source: https://github.com/hacksider/Deep-Live-Cam/releases/tag/2.7-beta