lamos vs Matrix LLM
Price, ratings, monetisation and update history for both apps, side by side β with what reviewers say about each.
lamos
Read full descriptionHide full description
Lamos is a private, on-device AI assistant for iPhone and iPad. Download a model once, then chat fully offline with no account, no cloud processing, no tracking and no limits. Analyze text, documents, and images with supported models, then tune settings for speed or quality. Your conversations stay on your device. Available models are: - Qwen 3.5 (Alibaba): text + image (vision) - Gemma 4 (Google): text + image (vision) - Ministral (MistralAI): text + image (vision) - Bielik (SpeakLeash): text - LFM 2.5 (Liquid AI): text - LFM 3 (Liquid AI): text Use built-in Assistants to improve your interactions with models: - General - Coding Helper - Document Summarizer - Brainstorm Partner Edit any Assistant or create your own. Good to know: At least 4 GB device RAM available - 6 GB or more (even 12 GB) recommended for best performance. App was tested on iPhone 15 and iPad Pro M1.
Matrix LLM
Read full descriptionHide full description
Matrix LLM β Private RAG for your documents Ask your own files and get answers with sources β right on your iPhone and iPad, offline and private. Researchers, analysts and journalists can work through confidential material on the go without uploading a single page. Matrix combines retrieval-augmented generation (RAG), on-device vector search and an interactive knowledge graph (ontology) in one app. Key features: β’ Local AI built in: language and search models run on the device. On first launch one tap sets them up β no account, no server, works offline. β’ Chat with your documents: ask across one or many collections. Every answer cites its sources, and each citation opens the exact passage. β’ Attach files to the chat: ask about a file right away, then keep it as a collection or discard it. β’ Questions in another language: better matches when your question is in a different language than the documents. β’ Older PDFs: update them once and every passage names its exact page. β’ Knowledge graph: people, organizations, places and their relationships are extracted automatically and shown as an interactive graph. Export to PowerPoint. β’ Tables: ask the same questions of every document and get a table with sources. Export to Excel. β’ Documents: PDF, Word, Excel and PowerPoint (also legacy formats), OpenDocument, e-books, e-mails with attachments, HTML, text and images including iPhone photos with text recognition (OCR) β even whole ZIP archives. Many Office formats, including older ones, are read reliably. β’ Share menu: send files from other apps to Matrix. β’ Spreadsheets: Excel and CSV files become queryable; the assistant calculates exactly with read-only SQL. β’ Hybrid search: semantic vector search and keyword search, optionally reranked. β’ Cloud optional: connect OpenAI, Azure OpenAI, Anthropic, Google Gemini, any OpenAI-compatible endpoint or an Ollama server on your network. β’ Transparent: the Prompt Inspector shows exactly what was sent to the model. β’ Privacy first: with local models your files never leave your device. For research, analysis, journalism, studying and everyone who keeps their documents confidential.
Screenshots
Verdict
The clearest difference is update cadence: Matrix LLM at weekly against lamos's monthly. On price, ads, in-app purchases and device support there is nothing between them.
Scored on Price Β· Rating Β· Positive reviews Β· Number of ratings Β· Update frequency Β· Ad data Β· In-app purchases Β· Monetization Β· Best chart rank Β· Devices Β· Requires iOS
Both are free to download. Neither carries in-app purchases, so what you see is what you pay.
lamos ships an update monthly, Matrix LLM weekly. Both last shipped on November 18, 2026.
| Parameter | lamos | Matrix LLM |
|---|---|---|
| Price | Free | Free |
| Update frequency | Monthly | Weekly β better |
| Ad data | No | No |
| In-app purchases | No | No |
| Devices | iPhone, iPad | iPhone, iPad, Mac β better |
| Requires iOS | 17.0 | 16.0 β better |
| Further details β not scored | ||
| Size | 16 MB | 80 MB |
| Age rating | 17+ | 4+ |
| Developer | MichaΕ Tadeusz Swierczewski | Matthias Nordwig |
In-app purchases
lamos
No in-app purchases
Matrix LLM
No in-app purchases









