CuriousLM is a free AI app that runs the model on your phone or computer, so your chats are written and kept there, encrypted. No account, no chat server. Words the model was unsure of stay underlined: tap one to see what else it nearly said, and take the answer another way. Voice and Pocket Wikipedia run on the device too, and on Android it can tell you who is waiting on you, with your permission.
The things you'd most want to ask an AI are the things you'd never want on someone else's server. So I built one that doesn't have a server.
CuriousLM runs the model on your own phone or laptop. Ask it why you keep apologising, paste in the message from your sister, let it read your calendar. It all stays on your device, encrypted. No account, nothing to sign up for, and it keeps working in airplane mode.
I'm Ebenezer, and this is the AI I wanted for myself.
A few things I'm proud of:
• Heat. Words arrive hot and cool to ink. The ones the model was unsure of stay underlined. Tap one to see what else it nearly said, and take the answer that way instead.
• It speaks first (Android). With your permission, source by source, it reads your notifications and calendar, then tells you who is waiting on a reply and what's due. Everything happens on your device, and nothing ever leaves it or goes to a server.
• Voice, Pocket Wikipedia and a journal, all on the device. Wikipedia's 121,797 most read articles fit in 60 MB, and answers cite them.
Honest limits: small models are slower and less clever than the giant ones in the cloud, and it's best on devices with 8 GB of RAM or more.
I'd love to know what you'd ask an AI if you knew nobody else could read it. And please tell me where it falls short.
does the 2B model hold up on a phone with less than 8 GB of ram, or does it just crawl?
The no-account part is what sells this for me. Most 'private AI' still starts with a sign-in and a server somewhere, so running the model on the device and keeping chats there, encrypted, actually matches the pitch. Pocket Wikipedia living on the device is a nice touch too — offline lookup without a round trip is exactly what local models are good at. Curious how the 2B model feels day to day compared to the cloud models people are used to — where do you notice the gap most?
I need to give this a try - I already have three AI apps on my phone, and none of them really work for me
the underlined alternatives look useful for exploring a different wording
i would keep that separate from fact checking since a model being sure of a word doesnt tell u whether the answer is correct
The underlined words are an interesting detail. They give someone a reason to pause over a sentence that otherwise sounds completely sure of itself.
About CuriousLM on Product Hunt
“Private AI that runs on your device. No account.”
CuriousLM launched on Product Hunt on November 2nd, 2026 and earned 93 upvotes and 11 comments, placing #15 on the daily leaderboard. CuriousLM is a free AI app that runs the model on your phone or computer, so your chats are written and kept there, encrypted. No account, no chat server. Words the model was unsure of stay underlined: tap one to see what else it nearly said, and take the answer another way. Voice and Pocket Wikipedia run on the device too, and on Android it can tell you who is waiting on you, with your permission.
CuriousLM was featured in Android (57.7k followers), Privacy (11.4k followers) and Artificial Intelligence (482.7k followers) on Product Hunt. Together, these topics include over 197k products, making this a competitive space to launch in.
Who hunted CuriousLM?
CuriousLM was hunted by Ebenezer Don. 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 CuriousLM stacked up against nearby launches in real time? Check out the live launch dashboard for upvote speed charts, proximity comparisons, and more analytics.
Ebenezer Don
The things you'd most want to ask an AI are the things you'd never want on someone else's server. So I built one that doesn't have a server.
CuriousLM runs the model on your own phone or laptop. Ask it why you keep apologising, paste in the message from your sister, let it read your calendar. It all stays on your device, encrypted. No account, nothing to sign up for, and it keeps working in airplane mode.
I'm Ebenezer, and this is the AI I wanted for myself.
A few things I'm proud of:
• Heat. Words arrive hot and cool to ink. The ones the model was unsure of stay underlined. Tap one to see what else it nearly said, and take the answer that way instead.
• It speaks first (Android). With your permission, source by source, it reads your notifications and calendar, then tells you who is waiting on a reply and what's due. Everything happens on your device, and nothing ever leaves it or goes to a server.
• Voice, Pocket Wikipedia and a journal, all on the device. Wikipedia's 121,797 most read articles fit in 60 MB, and answers cite them.
Honest limits: small models are slower and less clever than the giant ones in the cloud, and it's best on devices with 8 GB of RAM or more.
I'd love to know what you'd ask an AI if you knew nobody else could read it. And please tell me where it falls short.