This page is one real taste profile. Every score below is a prediction the model made before the film was rated.

Sample taste profile

Letterboxd remembers what you watched. palette predicts what you’ll love.

This page is one real taste profile — a year of ratings turned into a model that scores every film you haven’t seen. Yours will look like this.

Build your model See how the model works →

What's next

My next ten films, ranked by what I’ll actually think of them.

Each card is a prediction made before I’ve seen the film. The orange number is the model’s star rating in advance.

1972 · Federico Fellini

1973 · Martin Scorsese

1941 · Orson Welles

1927 · Fritz Lang

1972 · Andrei Tarkovsky

Click any card to open the film. The model breakdown page shows which of the ten components produced this score — see how the model works.

Honest predictions

Where the model gets me right, and where it gets me wrong.

Two films I’ve already rated. The model predicted one, I rated another. The gap is what teaches it.

↑ Loved it more than the model expected

↓ Liked it less than the model expected

What's different

A taste model that’s only yours.

Other apps recommend what people like you tend to like. palette predicts what you would actually rate it.

Different problem. Cleaner answer.

Ten components — genre and mood, director and cast, script direction, franchise patterns — get re-weighted by which ones predict your stars best. Your top weight may not be anyone else’s. See how the model works →

Private alpha

Want a model trained on your Letterboxd history? Drop me an email.

Get an invite Or read the build log →