Movie recommendations that start with you

Find something worth watching.

Pick a few movies you already love. I'll blend their taste patterns and look through 89.6K movies to make you a fresh shortlist.

01

Build your movie mix

One movie works. Three or more gives the model a much better read.

Or start with
0/5 picked
See how it works

Start anywhere

A favorite is all the model needs.

Pick one movie for a focused match, or mix a few together to find the overlap in your taste.

  1. 1 Pick up to five favorites
  2. 2 The model blends their learned signals
  3. 3 Save the movies that make the cut

A real model behind a simple screen

Not a hardcoded list in a nice jacket.

Shortlist trains on the same one-to-five favorites flow you use here, retrieves against the full catalog in Go, and adds TMDB details so the results feel like movies instead of database rows.

Movies searched
89,585
Learned profiles
199,378
Catalog reach
14.5%vs. 0.12% popularity
Recall@100
0.331vs. 0.228 popularity
Want to poke at the model?Try an anonymous MovieLens viewer profile.

Known profiles use their rating history. New IDs honestly fall back to popular starter picks.