25 supported formats
Documents, images, audio, and video in one mobile workspace.
PRESS KIT
Product facts, screenshots, and official links for writers, reviewers, and partners.
Contact the teamWHAT'S NEW
LocalRAG! 3.7.0 can process collection-wide questions across every document while showing progress and coverage. Local AI now adapts context and memory use to the selected model and device. New local models are LFM2.5 2.6B, LFM2.5 8B (MoE), and Gemma 4 E4B on iOS, plus LFM2.5 1.2B on Android. Model downloads resume in the background, and Android can continue answer processing outside the foreground. Cloud AI adds Claude Sonnet 5.5. The release also improves answer completeness checks, multilingual retrieval, and Markdown rendering. Documents remain on the device; in Claude mode, only the question and retrieved excerpts are sent to the API, not the original file.
LocalRAG! lets people ask natural-language questions about PDFs, Office files, EPUBs, images, audio, and video — then return to the page or timestamp behind an answer.
Documents, images, audio, and video in one mobile workspace.
Jump from an answer to the original page or audio/video timestamp.
Extraction, indexing, and search run on-device; local AI can also work offline.
iPhone, iPad, and Android.
English, Japanese, German, French, Spanish, Korean, Brazilian Portuguese, and Simplified Chinese.
When Claude is selected, LocalRAG! sends the question and relevant retrieved portions — not the entire source file.
Use the official name “LocalRAG!” and link readers to the product site or store listing.
