RestoraX

Overview

RestoraX is a self-hosted toolkit for repairing video and audio, aimed at people working with aging recordings and archival material. It is best suited to technically comfortable users who want to shape their own restoration workflows. Its breadth is a draw; it is not a plug-and-play desktop enhancer.

Key features

The project describes 25 models for tasks including upscaling, colorization, face restoration, interpolation, scratch and dust removal, deinterlacing, stabilization, SDR-to-HDR conversion, and audio restoration. That range makes RestoraX useful for collections with varied problems, rather than only footage that needs enlargement. Its free plan supports upscaling up to 4x, artifact removal, frame interpolation, and face recovery.

A visual pipeline builder joins processing steps in a directed acyclic graph, with typed ports, parallel branches, merge strategies, retry policies, and progress tracking for each branch. This gives users control over complex, repeatable workflows, but requires more planning than a single-purpose enhancer. Automatic model-weight downloads from HuggingFace Hub reduce one setup chore, while plugins distributed as separate PyPI packages let developers add third-party restorers.

Users can work through a React web interface, a FastAPI REST API with WebSocket progress, or a command-line interface. A ComfyUI node pack offers another option for people already using that environment. The documented stack uses Celery and Redis for background jobs, PostgreSQL or SQLite for data storage, and MinIO in its production Docker setup; that flexibility suits managed deployments better than users seeking a minimal installer.

Pricing

There is no paid tier or free trial. The free plan includes the core restoration capabilities, so the main cost is operating the software yourself rather than a subscription. The MIT license permits use, modification, and distribution subject to its terms.

Platforms

RestoraX supports web, API, Linux, and self-hosted use. Docker Compose setups are documented for development and production. Minimum requirements include Python 3.11, 8 GB of RAM, 5 GB of disk space, and FFmpeg. CPU processing works, but the project recommends CUDA 12.1 or later and at least 8 GB of GPU VRAM; users without suitable hardware should weigh the likely setup and processing demands before choosing it. The maker documents an MIT license, but no security certification or compliance standard.

Who it's for

RestoraX makes the most sense for technically capable users handling old films, home videos, archival footage, anime, VHS, or newsreels who want to combine multiple restoration steps and run them on their own infrastructure. Its range of models and interfaces is less compelling for someone who only needs a quick enhancement, a conventional desktop workflow, or a service that handles deployment for them.

Pros and cons

• Pro: The 25-model range covers video, image, and audio restoration tasks, useful when a collection has more than one kind of damage.
• Pro: The DAG builder and API, CLI, web UI, and ComfyUI options give workflow builders several ways to integrate and monitor processing.
• Pro: The software is free and MIT-licensed, with self-hosted CPU processing available.
• Con: Python, FFmpeg, storage, and memory requirements create a real setup burden compared with a packaged desktop app.
• Con: The recommended GPU setup calls for CUDA 12.1 or later and 8 GB or more of VRAM, which may rule out users without compatible hardware.

Alternatives

For other options, see AI Video Restoration Software. Vapourkit is a free, open-source Linux and Windows app, a better fit for users who want a conventional app on either desktop platform. Aiarty Video Enhancer suits macOS and Windows users who prefer a desktop product and can accept its free version's 120-second video limit, watermarked exports, and lack of batch export, or consider its paid plans.

Klarity is a free local and offline image/video restoration tool for Linux and Windows, better suited to users who prioritize offline desktop use over RestoraX's workflow interfaces. TensorPix offers web, API, and self-hosted access; its free plan has weekly login credits, 1 GB storage, and input capped at 720p, so it may suit users who want that service model and can work within those limits.

HitPaw VikPea is an alternative for users seeking broader device coverage, including mobile platforms. Refocused offers a local Linux, macOS, and Windows workflow with all models and full processing in its free trial, but watermarks output. Cognitech Video Investigator is a paid alternative for users considering a cloud video-investigation product; its listed annual plan is $1,399 a year. Intelion Video Lab is another option for Windows or self-hosted use, with on-premise or secure hybrid-cloud deployment described for its Analytics Bus module.

Verdict

Choose RestoraX if you want a free, self-hosted toolkit with a broad set of restoration models and the flexibility to build multi-step workflows. Its principal advantage is that combination of breadth and control without a software charge. Look elsewhere if you need a turnkey desktop enhancer or do not want to manage the stated software and hardware requirements.