NeuRelease

a fast neural parser for torrent and release names
GitHub
What the model read
Fields
Verdicts
Speed

One shared vCPU on Vercel's free tier, which is what this page costs to run and not what the parser costs. The same model reads a name in about 2,400 µs through the Python binding on a desktop Ryzen, and far less again from C++ in a batch, where the work is shared across names and threads. Round trip is network and JSON on top.

Response JSON