MLXHub: Local AI & LLM Server
Juan Colilla · Developer Tools
Independent analytics for this App Store listing. AppTracker is not affiliated with Juan Colilla.
Chart standing · US
What changed · US
Listing changes AppTracker has observed for this app in this storefront — new versions, price moves, rating shifts, and store copy.
Qwen3.6 35B-A3B, a model that until now would not fit on a phone, runs on your iPhone and your iPad. It gets there through NAND expert streaming: instead of holding the whole model in memory, it fetches each expert the moment that expert has something to say, and lets go of the ones that don't. Supported vision models also stop carrying weight you are not using. While you type, the vision side stays asleep and only wakes up when you attach a photo, so a text conversation takes noticeably less memory than it used to. The assistant can reach into your apps: your calendar, your reminders, your Health data, your photos or your music library, always with your permission, call by call. And whatever you explain once can be saved as a Skill, so you never have to explain it again. It has memory now, too. What is worth knowing about you carries over to your next conversations, and you decide what stays: review, pin or delete any memory from Settings. Distributed inference across several of your devices on the same local network is no longer an experiment. It is simply part of the app. Joining a mesh is still free. And your iPhone or iPad can serve the rest of your home: chat and embeddings for the apps and scripts on your local network, over the standard routes /v1/chat/completions and /v1/embeddings, with authentication and local discovery. Support depends on the model you pick and the device you have. Write tools, Skills, letting the assistant save memories on its own, hosting a mesh and the LAN server are part of MLXHub Plus, which is exactly what keeps all the rest coming. Thank you for being here.
Price & rating history · US
Reconstructed from the listing changes AppTracker has observed in this storefront. A field is charted once it has changed more than once, and each line starts at the value it moved from.
Average rating · ★ 4.25 now · 7 changes on record
Ratings breakdown · US
The star breakdown behind this app's rating in this storefront.
In-app purchases · US
What this app sells inside itself, grouped by kind and priced for this storefront.
Screenshots
Keywords this app targets
Search terms AppTracker associates with this app — estimates, not search volume. ❝ = in the app's title; a highlighted number is its current position for that term.
Similar apps
Competitors and close alternatives, as estimated by AppTracker.
Local LLM Server Pro
Kevin Tang
2 shared topics · 4 shared title words
LLM Server
Linosec
2 shared topics · 3 shared title words
Local LLM: Private Secure Chat
Pocket Bits LLC
1 shared topic · 2 shared title words
Think Local: AI Private Chat
Digital Hole Pvt. Ltd.
2 shared title words
Private LLM - Local AI Chat
Numen Technologies Limited
4 shared title words
Vanta: Local AI & Self-Hosted
EZEL BAYRAKTAR
4 shared title words
ChatLLM: AI Assistant & Chat
Abacus.AI, Inc.
1 shared topic · 1 shared title word
SoloLLM
Stephen Thorn
1 shared topic
About
Your local AI assistant for iPhone and iPad. Chat offline with downloaded models, use native tools and keep useful facts for future conversations. You choose the model and control its permissions. An assistant that can help you act Ask about your calendar, reminders, Health data, photos or music library using supported tools. Choose which tools to authorize. With MLXHub Plus, create events and reminders, write supported Health data and save reusable instructions as Skills. Tool support and results depend on the model you choose. Memory you control Keep useful facts for future chats. Review, pin or delete stored memories in Settings. With Plus, compatible models can save or remove facts through the memory tool. Changes apply from your next conversation. More room for capable models Explore and download language and vision models for your device. Supported MoE models can load experts from storage as needed, making larger models accessible without keeping every expert in memory. On supported vision models, text-only chats use less memory by loading vision components only when needed. Model capacity and performance depend on your device, model and context. Text, photos and documents Write, attach a photo or add a document and ask a compatible model to help you understand it. Set the assistant's personality or customize instructions for a conversation. Apple's on-device model is also available on supported devices with Apple Intelligence enabled. Your devices, working together Connect compatible devices on your local network to run supported models together. Joining a mesh is free; hosting requires MLXHub Plus. A local server for your apps Use your iPhone or iPad to serve chat and embeddings to applications and scripts on your local network. The server supports standard routes including /v1/chat/completions and /v1/embeddings, with authentication and local discovery. The optional LAN server is part of MLXHub Plus and runs while the app is in the foreground. All yours, unmetered No token limits, no message caps, no monthly quota, and no AI company keeping an eye on what you write. Inference happens inside your device, and your conversations are not uploaded to any cloud AI service: an assistant and a server you can use without limits, under your control. And if you would rather not subscribe, there is the one-time Lifetime purchase: unlimited inference for life and every feature that lands later, at no extra cost. Because this keeps getting better — with every update we ship, every week as new open AI models appear, and every year with each new device and iOS release Apple introduces. Privacy and connections Local inference does not require uploading your conversations to a cloud AI service. Downloaded models can work offline. Features you enable, such as web tools, the LAN server and mesh, communicate with their respective services or devices. No ads or required MLXHub account. Anonymous product analytics and diagnostics can be disabled in Settings. See our Privacy Policy for details. Terms of Use (EULA): https://www.apple.com/legal/internet-services/itunes/dev/stdeula/
What's new · 2.2.1
Qwen3.6 35B-A3B, a model that until now would not fit on a phone, runs on your iPhone and your iPad. It gets there through NAND expert streaming: instead of holding the whole model in memory, it fetches each expert the moment that expert has something to say, and lets go of the ones that don't. Supported vision models also stop carrying weight you are not using. While you type, the vision side stays asleep and only wakes up when you attach a photo, so a text conversation takes noticeably less memory than it used to. The assistant can reach into your apps: your calendar, your reminders, your Health data, your photos or your music library, always with your permission, call by call. And whatever you explain once can be saved as a Skill, so you never have to explain it again. It has memory now, too. What is worth knowing about you carries over to your next conversations, and you decide what stays: review, pin or delete any memory from Settings. Distributed inference across several of your devices on the same local network is no longer an experiment. It is simply part of the app. Joining a mesh is still free. And your iPhone or iPad can serve the rest of your home: chat and embeddings for the apps and scripts on your local network, over the standard routes /v1/chat/completions and /v1/embeddings, with authentication and local discovery. Support depends on the model you pick and the device you have. Write tools, Skills, letting the assistant save memories on its own, hosting a mesh and the LAN server are part of MLXHub Plus, which is exactly what keeps all the rest coming. Thank you for being here.
Details