AI Movie Recommender

Write a prompt describing your situation: films you have loved recently, the mood you are in, who you are watching with, or a specific kind of story you want. The AI reads your description and returns a list of 8-10 films or series. Each entry includes the title, year, director, genre, a rating, where it streams, and a sentence or two explaining what connects it to your request.

Everything from mainstream Hollywood to niche subgenres: slow-burn Scandinavian crime, New Wave French cinema, J-horror, mid-century Hollywood noir, Brazilian social realism, body horror, coming-of-age dramedies, heist films with unreliable narrators. Naming the sub-genre directly in your prompt gets you sharper results than using broad labels like "action" or "drama."

Treat the streaming info as a starting point, not a live lookup. Catalogs rotate constantly, and a title the AI places on Netflix may have since moved or left. Verify on the platform directly or check JustWatch before sitting down to watch. The film title and year in each result are reliable; the platform assignment is a reasonable guess from training data.

Yes. Add a decade range, a country of origin, preferred runtime, the audience you are watching with, or themes you want to avoid. You can also name films you actively disliked and say why: "I hated the pacing in Tenet, avoid anything similarly fragmented." That kind of negative feedback is just as useful to the AI as listing what you enjoyed.

Yes. Say you are open to subtitles, or ask specifically for films from a country or language, and the AI draws on world cinema. South Korean thrillers, Iranian drama, Romanian new wave, Argentine noir, Japanese animation outside the Ghibli catalog: these are all fair territory. You do not need to know the films; describing the themes or mood you want is enough.

Usually 10 to 20 seconds. The output includes 8-10 fully written entries with descriptions and streaming notes, so it takes a moment longer than a simple text completion. The wait is worth it given the amount of structured information in each result.

Yes. Specify what you want in the prompt: a limited series, an ongoing show, a documentary series, or a mix of films and series. If you do not specify, the AI may return a combination. Asking for one format explicitly keeps the list focused on what you actually want to watch.

Platform algorithms are built around your watch history and the content each service wants to promote. This tool has no catalog to push and no watch history to anchor on. You describe what you want in plain language, and it searches across all titles it knows, not just the ones a particular service happens to own or has licensed this month. The results tend to include films spread across multiple platforms and eras.

You can copy the full list to your clipboard or download it as a markdown file. Logged-in users can also return to their generation history to pull up any previous list. If you generated a themed list for a specific occasion, it will still be there when that occasion comes around again.

Name two or three films you actually loved and say specifically what worked: the pacing, the performances, the ending, the visual style, something the story made you feel. Then state what you want to avoid. A prompt like "I loved the first two hours of Once Upon a Time in Hollywood but found the third act too abrupt, I want something with a similar mood and period setting but a more satisfying close" will produce far better results than "something similar to Tarantino."