This product was not featured by Product Hunt yet. It will not be visible on their landing page and won't be ranked (cannot win product of the day regardless of upvotes).
Search 1.26 million movies and TV shows the way you'd describe them to a friend. Type the plot, an actor, a character, a studio, a mood, or just a few emoji. It finds the film before you finish typing, typos and all. No sign-up, no filters to click. Titles also carry up to three emojis, and there's a guessing game hidden behind the "don't click here" button.
Hey Product Hunt, I'm Vu from Mostly Right! 😊
I built Mojirama to see how far search can go when it sits on a good dataset. It covers 1,260,645 movies and TV shows going back to 1874, and you search it the way you would describe a film to a friend.
Things to try:
Things to try:
- a plot: boy befriends alien (40 ms)
- an actor: tom hanks (40 ms)
- a director: tarantino (2 ms)
- a character: darth vader (4 ms)
- a studio: a24 (2 ms)
- something similar: movies like interstellar (42 ms)
- filters said out loud: korean thrillers (5 ms)
- an original title: la casa de papel (10 ms)
- a typo: game of thornes (28 ms)
- emoji: 👽🚀 (1 ms)
How it works. There is no model in the loop while you type. Every keystroke runs a search in memory: a keyword index that corrects typos from the catalogue's own titles and names, and an embedding index for anything that reads like a description. That is why the first results arrive in a few milliseconds.
The query goes to Clef (https://www.cloudflare.com/resou...) half a second after you stop typing. Clef answers a few dozen multiple-choice questions about the query: what job each word does (topic, person, studio, filter, sort order, mood or filler), a yes or no for each genre, the decade, the language, movie or TV, and the order. Answers below 60% confidence are ignored. The rest become a search plan, which is a set of filters, an order and the topic words left to search. The planned results replace the first ones when they differ.
The dataset is 16 sources joined into one table: TMDB for the catalogue, cast, crew and ratings, with Wikidata and Wikipedia on top. Every title is one row with no duplicates and the dataset refreshes every week. We also used Jev (https://typesafe.ai/blog/introdu...) to give each title tags and up to three emojis (Clef wasn't available at the time), which is what emoji search and the guessing game behind the "don't click here" button run on.
It runs on Google Cloud Run. Each instance has 8 vCPUs and 32 GB and holds the whole catalogue and both indexes in memory.
The dataset is public if you want to build on it: https://mostlyright.md/datasets/...
Post any search it gets wrong, with what you expected to see.
My favourite title in the catalogue: search "passage of venus". It's a string of photographs of Venus crossing the sun, taken in 1874, which is where the catalogue starts.
After that, try to break the search and tell us what it got wrong.
About Mojirama on Product Hunt
“Fast movie search that understands what you mean”
Mojirama was submitted on Product Hunt and earned 4 upvotes and 2 comments, placing #101 on the daily leaderboard. Search 1.26 million movies and TV shows the way you'd describe them to a friend. Type the plot, an actor, a character, a studio, a mood, or just a few emoji. It finds the film before you finish typing, typos and all. No sign-up, no filters to click. Titles also carry up to three emojis, and there's a guessing game hidden behind the "don't click here" button.
Mojirama was featured in Emoji (11.7k followers), Movies (15.1k followers) and Search (18.1k followers) on Product Hunt. Together, these topics include over 7.9k products, making this a competitive space to launch in.
Who hunted Mojirama?
Mojirama was hunted by Vu Hoang Anh. A “hunter” on Product Hunt is the community member who submits a product to the platform — uploading the images, the link, and tagging the makers behind it. Hunters typically write the first comment explaining why a product is worth attention, and their followers are notified the moment they post. Around 79% of featured launches on Product Hunt are self-hunted by their makers, but a well-known hunter still acts as a signal of quality to the rest of the community. See the full all-time top hunters leaderboard to discover who is shaping the Product Hunt ecosystem.
Want to see how Mojirama stacked up against nearby launches in real time? Check out the live launch dashboard for upvote speed charts, proximity comparisons, and more analytics.
Vu Hoang Anh