Lorqa - Local AI Translator

About this app

<p>Lorqa is an AI-native translation and real-time subtitles app for iOS and macOS.</p>
<p>It makes text conversion focused and efficient—large language models run locally on your device to translate text between languages accurately and naturally. No internet is required, and there are no privacy concerns: everything happens on your computer. Start any video (or audio) in any app to begin low-latency, accurate transcription; the local LLM translates in real time, and the floating bilingual subtitles can be freely dragged and customized in multiple ways. It supports transcription in 25 languages, with multiple models to choose from, helping users understand foreign-language videos.</p>
<p>Key features:</p>
<p>Runs LLMs locally; supports Qwen3 and Gemma3; now also supports importing models directly from Ollama and LM Studio<br />
Works offline, even without an internet connection<br />
Data is stored only locally to keep your content secure<br />
Clean, distraction-free interface focused on translation and real-time subtitles</p>
<p>In-App Purchase:</p>
<p>Payment will be charged to your Apple ID account at the confirmation of purchase. Subscription automatically renews unless it is canceled at least 24 hours before the end of the current period.Your account will be charged for renewal within 24 hours prior to the end of the current period. You can manage and cancel your subscriptions by going to your account settings in the App Store after purchase.</p>
<p>About Auto-Renewal</p>
<p>Payment will be charged to your Apple ID account at confirmation of purchase. Subscription automatically renews unless it is cancelled at least 24 hours before the end of the current period. Your account will be charged for renewal within 24 hours prior to the end of the current period. You can manage and cancel your subscriptions by going to your account settings on the App Store after purchase.</p>
<p>Email<br />
[email protected]</p>
<p>Terms: https://jdi-app.notion.site/Terms-5a99ba5bf942491b8299c07221bb5f7b<br />
Privacy: https://jdi-app.notion.site/Privacy-Policy-cdb1be6b87f64c06b7dbea6ffbbad83f<br />
EULA: https://www.apple.com/legal/internet-services/itunes/dev/stdeula/</p>
<p>My Journey</p>
<p>Translation is a very common need, and a variety of translation apps keep appearing on the market. Many large companies have entered this space, so there are numerous translation solutions. For a long time I used one particular translation tool daily — it was convenient: select text, press a hotkey, and the translation was done. But there were small annoyances during use, such as slow translation speed, limits on the amount of text that could be translated, and the need to use other methods for very long texts.</p>
<p>I’ve noticed that LLMs (large language models) are often used as translation tools now. LLMs have strong multilingual capabilities, and smaller models like Qwen3 and Gemma3 also provide very good translation quality. When using LLMs for translation there are many options: most are web-based, while some models are deployed locally. These products generally haven’t delivered a true end-to-end translation experience tailored to the use case — they often fall short.</p>
<p>I previously worked on PocketLM (a mobile LLM local inference client), so I have a good understanding of the capabilities of small LLMs and am familiar with inference and load times across different devices. That prompted me to start building an end-to-end translation app using LLMs. The implementation went fairly smoothly; in testing I found that Qwen3 0.6B 4-bit already performs on par with DeepL’s translation ability. I also added Qwen3 1.7B and the Gemma3 series as alternatives to give users multiple choices.</p>
<p>My motivation for making this product is quite personal. I’m curious whether this need exists for others as well, and I’d love to hear feedback on future plans.</p>