Subscribe
Sign in
Ollie
p/lora-2Private, Grown-up AI chat
Start new thread
Seunghwan

Lora - Local LLM for Mobile: Fast, Secure with Zero-Cost.

by
Lora is a local LLM for mobile, with performance comparable to GPT-4o-mini. It offers seamless SDK integration, guarantees complete privacy without data logging, and supports airplane mode. Try our app and build your own Lora-powered app.

Add a comment

Replies

Best
Seunghwan
Maker
📌
👋 Hello, I'm Seunghwan We're excited to introduce our new product, “Lora”. ☀️ What is Lora? Lora is a local LLM for mobile. It offers an SDK that lets you build a Lora-powered app with just one line of code. ✨ Who Should Use Lora? 🛠️ Developers who have built LLM-based AI apps but struggle with high operating costs 🌍 PMs preparing for global launches but facing obstacles due to data regulations 📡 Anyone who needs to offer AI services where internet connectivity is limited 🔎 Key Features of Lora ✅ GPT-4o-mini speed ✅ Simple SDK integration ✅ 100% private, no data saved ✅ Airplane mode support 🔥 Try Lora-powered AI app for free! We also launched a private AI assistant "Lora", Lora-powered AI chatbot app. It's 100% free to use. If you are interested in using it for development, please feel free to contact us. We can provide you with the SDK. A free plan is available, so don't hesitate to reach out. 🚀 Thrilled to Launch on Product Hunt! We’d love your feedback to make Lora a “Wow!” product. Questions or suggestions? DM me anytime. Thank you!
Woobeen Back
Maker
@seungwhan @sonu_goswami2 Thank you for your support! 😎😎😎
Hansol Nam
Maker
@sonu_goswami2 Thank you so much! 😊 Your support means a lot! 🚀✨
Junghwan Seo
Maker
@sonu_goswami2 SUPER THANK YOU ☺️
Seunghwan
Maker
@sonu_goswami2 Thank you for your support! We value real user feedback, so if you try it out and share your thoughts, we'd really appreciate it. Thanks again, and have a great day!
Woobeen Back
Maker
@seungwhan @sonu_goswami2 Hello! If you are okay, I'm here to answer :) When we lightweighted Lora, we used NAMMs which Sakana AI recently published(Cetin, Edoardo, et al. "An Evolved Universal Transformer Memory." arXiv preprint arXiv:2410.13166 (2024).), and a few more techniques from Peekaboolabs to lessen memory usage and unnecessary calculations Which leads to battery efficiency :) Now, Lora only consumes the same amount of battery as playing video games!
Evak Chan
Privacy and data protection are my top concerns when using AI tools. I appreciate the 100% private and no data saved feature of Lora and curious about how do you ensure 100% privacy?
Woobeen Back
Maker
@evakk Hello! Thank you for reaching out, and I really appreciate your questions about privacy :) Lora doesn't collect ANY private data except your login and does not collect your chat data(You can check that by enabling airplane mode!). Your chat data will be saved in your private storage, which no one can access. If you want to delete them, you can permanently delete them by flushing them away or deleting Lora. And after downloading the model, Lora does not require any internet access for evermore. So, Lora does not collect any data, and it is designed unable to collect any data. If you use Lora, you can talk about important things free from threat of leakage or hacking. Thank you!
Hansol Nam
Maker
@evakk First of all, thank you for your deep interest in Lora! We’ve conducted interviews and found that many users are concerned about sensitive or legal information being sent to servers. By processing everything solely on the user’s mobile device, we ensure that customers can confidently utilize AI features without any privacy concerns!
Junghwan Seo
Maker
@evakk If you install the app at first, just sign in with Google, put it in airplane mode, and use it until you delete it, you'll see what I mean 🤣
Seunghwan
Maker
@evakk The AI is installed directly on the user's mobile, ensuring that no data is transmitted externally. Additionally, we do not perform any data logging, guaranteeing 100% privacy.
Evak Chan
@peekabooooo I will give it a try 😃
Michael Vavilov
@seungwhan congrats on the launch!
Seunghwan
Maker
@michael_vavilov Thank you so much. If you're okay with it, would you be willing to try our product and share your feedback?
Woobeen Back
Maker
@seungwhan @michael_vavilov Thanks :) Using Lora SDK, you can build tons of AI products and more! Hope you like it
Hansol Nam
Maker
@michael_vavilov Thank you so much! 😊 Your support truly means a lot and gives us a huge boost! 🚀💙
Junghwan Seo
Maker
@michael_vavilov SUPER THANK YOU 🫡
Muhammad Waseem Panhwar
@hansol_nam Really impressed by Lora's focus on privacy and offline functionality – exactly what mobile AI needs. Excited to see how it evolves! Congratulations Team
Hansol Nam
Maker
@waseem_panhwer Thank you so much for your kind words! 😊 We truly believe that privacy and offline functionality are crucial for the future of mobile AI. Excited to keep improving and can’t wait to share more updates with you! 🚀🙌
Woobeen Back
Maker
@waseem_panhwer Thank you for your support! And I really appreciate that you found the true value of mobile AI. It is time to get free from server-based AI, and Lora is one of the keys to that :)
Junghwan Seo
Maker
@waseem_panhwer Thank you so much for your kind words! 🙌✨ We’re thrilled that you appreciate Lora’s privacy-first, offline approach. Once you give it a spin, we’d love to hear your honest thoughts—good or bad. 🤗🔒 Your feedback helps us keep evolving!
Changhoi Kim
wow, this is insane! it's like something out of a sci-fi. What is difference between the free tier and the paid?
Woobeen Back
Maker
@changhoi Hello! Thank you for your interest. Our demo app supports free & unlimited chat. but if you want to build one for yourself, you can use Lora SDK, which is divided into free tier and paid. Actually, every development tool is supported in the free tier. But if you want to make your on-device LLM service without any limits and a comfortable performance analytics dashboard, a paid plan will be the best option!
Junghwan Seo
Maker
@changhoi Lora is free to use, with a maximum of 20000 tokens per day, and paid usage is charged to companies that develop AI services using the our SDK 👷🏻‍♂️👷🏻‍♀️
Hansol Nam
Maker
@changhoi Glad you’re excited about it! 🚀 The free tier includes: • Single Request Limit: 2,000 tokens • Daily Request Limit: 20,000 tokens • Basic support With the paid version, you get: ✅ Unlimited Lora LLM Access – No token limits, so you can generate as much as you need. 📊 Performance Analytics Dashboard – Track and analyze AI performance across different devices. 🔧 Technical Support – Get direct assistance for development and implementation. In short, the paid version provides unlimited token usage, device-specific performance insights, and dedicated technical support to optimize your experience!
Seunghwan
Maker
@changhoi Try free tier first :)
Shiva Bajpai
@changhoi @hansol_nam Congratulations for the lauch @hansol_nam 🔥 It's too good & much needed thanks for building it 👍🏻
M S Sulthan
Great tool! Looking forward to upcoming features.
Seunghwan
Maker
@mssulthan We are preparing awesome features... You can look forward to it!
Hansol Nam
Maker
@mssulthan We’ll keep working hard to add even more great features! 😊 Truly appreciate your support—thank you so much! 🚀🔥
Woobeen Back
Maker
@mssulthan Thank you! We are also working on more features such as datapack for RAG; which allows to eliminate hallucination and add accurate informations. Please stay tuned! I can't wait to show you Lora-with better features!
Junghwan Seo
Maker
@mssulthan Stay tuned, and we'll make good progress so that one day you can say, “Huh? I know these guys!” ☺️
Michael Talreja
@seungwhan @hansol_nam @peekabooooo @woobeen_back Congratulations on the launch of Lora. I personally love the privacy aspect of this which is now a huge concern and a compulsion for any enterprise. Going to try this out right away and will spread the word with my colleagues. Keeping my eyes peeled on how this product grows:)
Woobeen Back
Maker
@seungwhan @hansol_nam @peekabooooo @michael_talreja Thank you so much for the kind words! We're thrilled that you appreciate the privacy features of Lora—it’s a top priority for us, especially in today’s privacy-conscious world. We're confident that Lora will not only provide the security you need but also offer flexibility and power at your fingertips. We’re excited to have you try it out and can’t wait to hear your thoughts as you explore its potential. Thanks for spreading the word—together, we can make this product even better!
Hansol Nam
Maker
@michael_talreja Thank you so much for your support! 😊 Privacy is indeed a major concern, and we’re committed to making Lora a truly secure and enterprise-friendly solution. 🚀 Excited for you to try it out—really appreciate you spreading the word! Looking forward to sharing our progress with you. 🙌🔥
Junghwan Seo
Maker
@michael_talreja We'll be working on it and so you can say somewhere, “Huh? I knew this guys!” Thank you :)
Shivam Singh
@seungwhan, such an awesome product this is ++ Running AI locally without privacy concerns or internet dependency is a big deal. The SDK approach makes it easy for developers to integrate, and I can see this being really useful for global apps facing data regulations. Solid work on this! Sending wins to the team :)
Seunghwan
Maker
@whatshivamdo You’ve truly grasped the essence of our product. We’ll keep improving to create even better ones, so stay tuned. It would be incredibly helpful if you could try it out and share your feedback!
Hansol Nam
Maker
@whatshivamdo One of our customers recently experienced a service disruption due to an internet outage caused by a hurricane. They even faced issues during a CES demo. I truly believe network-free AI is a crucial technology! Thank you so much for your warm support! 😊🚀
Junghwan Seo
Maker
@whatshivamdo SUPER THANKS ☺️
Tanmay Parekh
All the best for the launch @seungwhan & team!
Woobeen Back
Maker
@parekh_tanmay Thank you! Lora will allow everyone anytime and anywhere to use LLM. Welcome to the world of on-device
Hansol Nam
Maker
@parekh_tanmay Thank you so much for your kind words and support! 😊 Your encouragement truly means a lot to us and gives our team a huge boost on this special day. 🚀 We’ll keep pushing forward—really appreciate you being part of our journey! 🙌
Seunghwan
Maker
@parekh_tanmay Thank you so much. We're building something truly outstanding. Stay tuned!
Junghwan Seo
Maker
@seungwhan SUPER THANKS ☺️
Danilo Fiumi
Congrats for the launch 🎉🎉 I love the idea of complete privacy without data logging. Great job!
Woobeen Back
Maker
@danilofiumi Thank you for reaching out! We think one of the most important value of on-device AI is privacy and security. It is so good to hear
Hansol Nam
Maker
@danilofiumi Thank you so much! 😊🎉 Privacy is at the core of what we do, and we’re thrilled that you love the concept! 🚀 Your support means a lot—excited for you to try it out! 🙌
Junghwan Seo
Maker
@danilofiumi SUPER THANK YOU for your support 😊
Sungman Cho
I believe the future for on-device AI will coming soon :) Hope your progress bring the future forward!
Woobeen Back
Maker
@sungman_cho1 Thank you for reaching out! Lora will be a great tool for everyone that wants to build LLM service without any costs. Try our free & unlimited demo if you are interested :)
Hansol Nam
Maker
@sungman_cho1 the future for on-device AI is coming!!! 🚀 Thank you so much for your warm support—it truly means a lot! 😊
Junghwan Seo
Maker
@sungman_cho1 Thanks for the support! If you can, give it a try and experience the future of on-device 😆
Seunghwan
Maker
@sungman_cho1 So nice! We believe too. Thanks for your kind word!
Hanzala Siddiqui
Great addition to the AI world 👏🏽
Woobeen Back
Maker
@hanzala_siddiqui7 Thank you! With Lora, everyone can now experience the power of LLM anytime and anywhere, without limitations. It opens the door to a new world of on-device technology, making advanced AI more accessible and bringing innovative possibilities right to your fingertips. Welcome to this exciting new era!
Hansol Nam
Maker
@hanzala_siddiqui7 It really feels like on-device AI is becoming a part of our daily lives! 🚀 Thank you so much for your support—it truly means a lot! 😊
Junghwan Seo
Maker
@hanzala_siddiqui7 We'll keep improving to deliver more value to everyone! Please support us ☺️
Seunghwan
Maker
@hanzala_siddiqui7 I'm glad you think so! What you say aligns our mission statement, which is to empower AI everywhere!
Adam York
Congrats on the launch, @seungwhan! Offering a local LLM with easy SDK integration sounds incredibly useful—how does Lora handle updates and improvements without relying on cloud-based processing?
Woobeen Back
Maker
@seungwhan @adamyork Hello! Thank you for good question! Lora is designed to ensure seamless updates and continuous improvements while maintaining an on-device experience. Updates are delivered efficiently through lightweight patches, allowing users to keep their models up-to-date without depending on cloud-based processing. This approach not only preserves privacy and minimizes latency but also ensures that developers can easily integrate improvements via the SDK without interrupting workflows.
Hansol Nam
Maker
@adamyork Great question—thank you! 😊 We’re continuously improving the model and will enable dynamic downloads for updates. This way, users can leverage Lora’s local LLM on-device while still receiving the latest LLM improvements through patches, all without relying on cloud-based processing. 🚀
Seunghwan
Maker
@adamyork @woobeen_back Our team member wrote in detail. If you want to use, please contact us.
Junghwan Seo
Maker
@adamyork For more information, see Team members 🫡
Aleksandr Milenovich
Interesting! Goodluck with the launch @seungwhan!
Woobeen Back
Maker
@seungwhan @ficus_26 Thank you for your interest and support 🤗 You can use our demo app right now!
Hansol Nam
Maker
@ficus_26 Thank you so much for your support! 😊 It truly means a lot to us and gives our team a huge boost! 🚀🔥
Seunghwan
Maker
@ficus_26 Thank you so much.
Junghwan Seo
Maker
@ficus_26 Thank you for your comment 😊
Sanguk Park
Interesting! As this is local LLM can I use it without internet? if so it must be so helpful. Few weeks ago I visited Yosemite National Park, and the signal was unavailable. It must be so helpful if I can talk with AI in those situation.
Junghwan Seo
Maker
@psycoder You can also use it when you go into space or scuba diving, so try it out there and let us know 🤣
Seunghwan
Maker
@psycoder It's so true! That's one of the reasons why a local LLM is needed.
Hansol Nam
Maker
@psycoder It must have been really inconvenient due to the unstable internet. I will work hard on research and development to ensure that services can work without problems even in wide areas with unstable internet. Thank you so much for your warm encouragement :)
Chiara Maria Cervetta
Wow, congratulations to the launch! 🥳 🚀 🎉 I'm an enthusiastic consumer of conversational AI services and am excited to try the private AI assistant "Lora" app and also to share feedback soon! 🥳 ✨ Thank you very much! 😊 I wish you a lot of success and all the best! 🤩 🍀 🎉
Seunghwan
Maker
@chiaracervetta Thank you so much for your kind words and enthusiasm! We’re thrilled that you’re excited to try Lora and truly appreciate your support. Looking forward to your feedback! Wishing you all the best as well, and thanks again for your encouragement!
Hansol Nam
Maker
@chiaracervetta Since you love conversational AI services, you must be a true tech early adopter! 😊 Your feedback will be incredibly valuable for us. Thank you so much for your kind and encouraging comment! 🚀🎉
Chiara Maria Cervetta
@seungwhan, you're most welcome! 🥳 💐 Thank you so much for your kindness and I'm looking forward to giving feedback about the Lora chat AI soon. 😊 🙌🏻 I wish you and also the entire team all the best too and I'm so grateful to you all! 🥳 🍀 💐
Chiara Maria Cervetta
@hansol_nam, you're most welcome and I'm incredibly grateful for your kindness! 😊 🥳 💐 I'm looking forward to giving feedback about the Lora chat AI soon. 🤩 🙌🏻 I wish you and also the entire team all the best and I'm so grateful to you all! 😊 🍀 💐
Jeongmin Park
Really excited for this! As someone who’s always been mindful of privacy and security, this approach feels like a game changer! 🔥 wishing you great success with the launch!!! 🚀👏
Seunghwan
Maker
@jeongmin_park1 Thanks a lot! I’d love to hear your feedback on the app soon.
Hansol Nam
Maker
@jeongmin_park1 Sharing sensitive or confidential information on a server-based LLM can definitely feel burdensome. Thank you for your kind and warm comment! 😊
Isaac Martin Otim
This is amazing! What are the minimum mobile hardware requirements for this to work on device?
Woobeen Back
Maker
@izakotim Thank you!! In our test, ANY device with 8GB ram can run Lora(For example, Galaxy Note 9 or OnePlus 5). It is just matter of how long it takes to run inference :) In our best test device iphone 15 pro, it takes only 1.2 second to start inference!
Hansol Nam
Maker
@izakotim Thank you so much for your deep interest in asking about the minimum spec devices! 😊 We’re continuously optimizing our model to ensure the service works across a wide range of devices. 🚀
Seunghwan
Maker
@izakotim We believe that hardware performance will increase exponentially in the future.
Junghwan Seo
Maker
@izakotim If you want a smaller spec... ask to @hansol_nam and @woobeen_back 🤣
Zeshan Ather
@seungwhan, Many Congrats on your launch.
Seunghwan
Maker
@zeshan_ather Thank you for your support. I’d love to hear your feedback on the app soon.
Hansol Nam
Maker
@zeshan_ather Thank you for your kind and warm comment! 😊✨ It really means a lot! 🚀🎉
Junghwan Seo
Maker
@zeshan_ather Thank you for your support ☺️
Jan Ahrend
All the best for the launch! What’s planned next for Lora? 🎉
Seunghwan
Maker
@jmarten Thank you for your support! Our next plan is to find a PMF and make a revenue stream.
Hansol Nam
Maker
@jmarten Thank you so much for your kind words! 🎉 Your support truly means a lot! 😊 Next, we’re focusing on making Lora even faster, lighter, and more versatile while keeping privacy at its core. Excited for what’s ahead—stay tuned! 🚀🙌
Junghwan Seo
Maker
@jmarten We are now prioritizing listening to our customers to understand what they truly need 😘