FastVSR makes low-res, low-FPS video sharp and smooth: up to 4x resolution and 2x or 4x frame rate on dedicated GPUs. Use the web app or plug the REST API into your pipeline. Exact cost shown upfront, full refund if a job fails on our side.
Hi Product Hunt 👋
We built FastVSR because a lot of good footage is stuck at low resolution and low frame rate, and fixing it usually means a powerful local GPU, hours of setup, and guessing what the job will cost. We wanted it to be one upload, or one API call.
**How it works**
1. **Upload** your video (MP4, MOV, AVI, MKV or WebM). You see the exact credit cost before you confirm.
2. **Process:** a dedicated GPU server spins up just for your job, with no shared queue. Watch progress live.
3. **Download** the result, with your original audio intact.
**What you get**
- 🔍 Up to 4x resolution upscaling (up to 4K resolution)
- 🎞️ 2x or 4x frame interpolation for smoother motion
- 🎧 Your original audio track, preserved
- 📐 Transparent pricing from a published formula, so you know the cost before you commit
- 🛡️ Fair billing: full refund if a job fails on our side, and if you cancel mid-job you pay only for the GPU time already used
**Built for developers too** 🧑💻
Everything you can do in the app, you can do over our REST API. Create an API key in your dashboard, then:
- Estimate the credit cost of a job before you submit it
- Upload large files straight to storage with a presigned URL
- Submit jobs with the upscale and interpolation settings you want
- Follow progress live over Server-Sent Events, or poll the job status
- Download results, list and cancel jobs, and check your credit balance
The docs have copy-paste examples in cURL, Python and Go, so batch processing, CI/CD pipelines, or your own video tool built on FastVSR are a few lines of code away.
Under the hood, upscaling runs on a diffusion-based video super-resolution model, and cutting-edge AI-based interpolation.
Sign up to get free credits, no credit card required. We'd love to hear what you upscale first, and what you think of the quality and the API!