How to get AI recommendations from a mood or a film you loved
Tell the extension your mood or name a film you loved — "a slow-burn thriller like Zodiac" — and it returns suggestions. The query goes to Google Gemini along with a compact summary of your top genres, never your raw viewing history, and the returned titles are then scored against your local taste profile before you see them.
Takes about 2 min
Steps
Build a taste profile first
Rate or import at least five titles, or the scoring step has nothing to work with.
Open the recommendation feature
In the extension.
Describe what you want
A mood, a genre, or a specific film to use as a reference point. Plain language works — you're not picking from a menu.
Read the results with Taste Match in view
Suggestions are scored against your profile, so the percentage tells you how well each fits you specifically.
Worth knowing
- What's sent to the model: your query plus a compact summary of your top genres. What isn't: your raw viewing history.
- This is the one feature that makes a network request by design — it's a call to a language model. Everything else in the extension is local.
- Recommendation quality scales with how many titles you've rated. Below five, it falls back to globally popular suggestions.
- It's free for personal use, with no rate limits and no premium tier.
Questions
What information is sent to the AI?
Your recommendation query and a compact summary of your top genres — never your raw viewing history. The model returns titles, which the extension then scores against your local taste profile before displaying them.
How good are the recommendations?
Quality scales with how many titles you've rated. After around twenty they should feel meaningfully personalised; below five the extension falls back to generally popular suggestions rather than guessing.