Vibe
Describe a mood, a moment, a kind of night. Perfery reads tone, era, and craft — not just genre tags — and reaches past the obvious picks.
FOR FILMS YOU HAVEN'T SEEN YET
Describe a mood, paste a Letterboxd list, or pick a few films you love — Perfery finds what your habits have been missing.





Describe a mood, a moment, a kind of night. Perfery reads tone, era, and craft — not just genre tags — and reaches past the obvious picks.
Paste any public list or profile. Perfery reads every title as a whole and surfaces films the list curator almost certainly missed.
Add a few films you love — even a mix that doesn’t share a theme. Perfery finds what lives in the same world: similar mood, craft, and sensibility.
A mood, a list URL, a handful of films you love. No tags. No categories. However you enter, Perfery meets you there.
We read tone, era, craft, and the company a film keeps — then push past the obvious picks toward what your habit has been missing.
Swap any title. Hide what you’ve seen. Slide from crowd favorites toward deep cuts. Export a clean list to Letterboxd when the set feels right.
Everything you need to know about Perfery.
Perfery is a free web app that suggests films to watch next from a public Letterboxd list, hand-picked seed titles, or a short plain-language prompt—mood and tone, sure, but also references, comparisons, questions, or a specific idea.
It builds an editable grid of recommendations you can trim, swap, filter, and export as a CSV for Letterboxd.
Yes. Perfery is free to use in your browser. We do not charge for recommendations, filters, or CSV export.
You choose an input: paste a public Letterboxd list or profile URL, search and add any films as seeds, or type a free-form ask—anything from “slow rainy noir” to “films like X” or “what fits a Sunday with friends.” Perfery uses TMDb metadata plus keyword and similarity-style matching to propose a related set of films.
You stay in control: adjust filters, remove rows, swap alternates, then export if you want a Letterboxd-ready list.
No. Letterboxd is optional. You can run suggestions from a public list URL, from films you pick in the app, or from a free-form text prompt without ever signing in.
To hide films you have already seen, use Import Letterboxd in the nav and upload a Letterboxd data export (CSV or ZIP) so Perfery can match against your diary—only in this browser.
For list-based runs, paste any public Letterboxd list URL, your films page (letterboxd.com/username/films/), or a rated slice (for example letterboxd.com/username/films/rated/4-5/). Perfery reads those public pages to build input for recommendations.
For diary-based “hide watched,” use Import Letterboxd in the nav to upload a Letterboxd export file (CSV or ZIP). Your export is processed in the browser; nothing is posted to Letterboxd.
Use public list URLs, your profile films page, or a public rated page. Private lists must be set to public on Letterboxd so the page can be read.
Perfery only reads public Letterboxd pages. If a list is private, the site cannot see its titles—make the list public or paste a different public URL.
Vibe mode is free-form text instead of starting from a list. You can lean on mood, tone, and setting—or name a reference film, ask a question, sketch a plot or era, or say “something in the vein of…” in one short paragraph.
Perfery interprets that text together with TMDb data to suggest films that fit what you wrote.
Yes, in a narrow way. Perfery uses AI mainly to translate your plain-language prompt (whether that is mood, a comparison, or a question) into structured hints—things like genres, themes, or candidate seeds—so the rest of the pipeline knows what to look for.
The actual recommendations are ranked by Perfery’s custom-built algorithm on top of TMDb data: keyword and similarity-style matching, list-based signals when you start from Letterboxd or hand-picked films, filters, and niche lean—not by a chat model choosing or ordering your grid.
When you turn on Letterboxd-based exclusion and Perfery has your diary from an imported export, it can leave out titles you have already marked as watched so suggestions skew toward unseen films.
Nothing is posted back to Letterboxd; it is only used to filter your results in Perfery.
Your exclusion preference and diary data used for “hide watched” (from a Letterboxd export you import) are stored locally in your browser (localStorage), not on our servers long term.
When you request recommendations, your list URL, seed titles, or prompt text is sent to our servers to run that request and is not retained after the request finishes. See our Privacy Policy for full detail.
Yes. After you have a results grid, you can narrow films by streaming availability using provider data from TMDb (region and catalog coverage depend on what TMDb reports).
IMDb ratings are fetched from imdbapi.dev when available. Metascores come from Metacritic and are cached for faster repeat lookups. Both are shown for convenience and may not always be present for every title.
Yes. Download the CSV from your grid and import it into Letterboxd using their list import flow.
No. Perfery is an independent project. It is not affiliated with, endorsed by, or operated by Letterboxd or TMDb. Film metadata is powered by TMDb; Letterboxd pages or exports are only used when you choose those inputs.