# mycellm — Run LLMs — Your node for distributed AI > Your node for distributed AI (iPhone/iPad app by Dodecki Labs.) - Source: https://appshunter.io/ios/app/mycellm-run-llms/id6761091607 (this page in markdown: same URL + `.md`) - Developer: [Dodecki Labs](https://appshunter.io/developer/1571436485) - Category: Utilities, Developer Tools - Price: Free - Age rating: 12+ - Requires: iOS 17.0 · 42 MB - Languages: American English - Released: 2026-06-24 - Data updated: 2026-09-04 - User reviews in markdown: https://appshunter.io/ios/app/mycellm-run-llms/id6761091607/reviews.md ## What is mycellm? Mycellm turns your iPhone or iPad into a node on a distributed AI inference network. Run open LLMs on your own hardware — no account, no telemetry, no cloud middleman. Use it three ways, and own every layer. LOCAL — ON YOUR DEVICE Load a model and run inference entirely on-device. Fully offline, airplane-mode friendly. On-device Privacy Guard scans prompts for secrets before anything hits the wire. TRUSTED DEVICES — YOUR OWN NETWORK Point your phone at a node you trust — your Mac Studio, your iPad Pro, your homelab — and run models far bigger than any single device could hold. Your phone stays light; a machine you own does the heavy lifting. No cloud. PUBLIC — THE FREE NETWORK Tap the public mycellm net: open models served by GPUs worldwide. Earn credits by serving spare compute, spend them to run larger models. Peer-to-peer, no subscription. FEDERATE Stand up your own network with its own trust policy and credit ledger, mint invite tokens, and bridge trusted networks with cross-network routing. You decide membership, models, and data flow. UNLEASH THE iPAD PRO An iPad Pro (M4/M5) is a serious AI machine that sits idle most of the day. Put it to work: more unified memory runs larger models, Metal delivers 30+ tok/s on multi-billion-parameter models. Serve your iPhone, your team, or the network — first-class on any device that runs a model. FOR PEOPLE WHO RUN THEIR OWN STACK - Serves the same chat-completions REST API your tools already speak — point curl, Claude Code, or your own clients at your device. - Vision models: attach an image and ask, on-device. - Live dashboard: tok/s, credits, peers, node health. - Auto hardware-fit: picks quantizations that match your RAM. - Open source, Apache 2.0. Audit it, fork it, run it. No black boxes, no lock-in. It's BitTorrent for LLMs — every node both uses the network and makes it stronger. Full docs and use cases at mycellm.dev. Bring your own models, keep your data, and run AI on your terms. ## Version history (last 5 releases) ### 1.3.0 — 2026-08-25 Pick your model, right from the chat. • MODEL & QUALITY PICKER — Choose Automatic (still the default), a quality floor like "Capable (13B+)", or one specific model. Live counts show what each tier can actually reach right now, and a tier nothing meets says so instead of quietly giving you something smaller. • JOB QUEUE — Send work to your network and let it run when a device is free: an iPad on the charger tonight, a Mac that wakes in the morning. Nothing is lost if you close the app. Every waiting job explains what it's waiting for, in plain language. • SWARM PROGRESS — When several models answer together, you can see it happening: "Asking 3 models on aurora, hokulea…", then "Synthesising 3 answers…" — instead of a dot animation that looks the same as a hang. • BETTER FLEET AWARENESS — This node now tells your other devices what it can genuinely do: which models can only embed rather than chat, and whether it's on battery, thermally throttled or on an expensive network. Work gets planned around a phone that's in no state to serve it. • ACCURATE TOKEN COUNTS from relayed and multimodal models, which previously reported zero. Requires a mycellm node running 0.8.0 for the queue and quality tiers. Everything else works with any recent version. ### 1.2.0 — 2026-08-17 Reliability and refinement across the app. • More dependable model downloads, including on metered connections • Clearer status in Chat when a model isn't ready to use • Improved visibility into what your node and network are doing • Assorted fixes and performance work under the hood ### 1.1.0 — 2026-08-07 New in 1.1.0: - Tip jar: support mycellm development with a one-time tip (Settings) - Core protocol updated to 0.6.3 - full parity with the mycellm network - Model downloads are verified against published checksums - Streamed replies no longer show chat-template markers - Bounded-memory option for long conversations on smaller devices - Reliability and performance fixes ### 1.0.1 — 2026-07-02 Multi-network support and smarter memory management. • Join multiple networks at once — stay on the public network while also joining private networks you host • Redesigned Network tab: per-network cards with live status, endpoint settings, and Enable/Disable • Custom bootstrap host and port with one-tap Reconnect • Local-network permission is now requested properly, so private coordinators on your LAN connect reliably • Fleet administration: check status and manage your nodes remotely • Smarter memory check before loading models — clamps context length instead of running out of memory • Accepts max_completion_tokens for chat-completions compatibility • Stability improvements ### 1.0.0 — 2026-06-25 No release notes. ## More apps by Dodecki Labs - [Ensembl: Get it together](https://appshunter.io/ios/app/ensembl-get-it-together/id1572533173) - [WunderGuide: Travel Concierge](https://appshunter.io/ios/app/wunderguide-travel-concierge/id6745361263) - [GentlePrep: Colonoscopy Prep](https://appshunter.io/ios/app/gentleprep-colonoscopy-prep/id6760452252) - [Span: Physics Bridge Builder](https://appshunter.io/ios/app/span-physics-bridge-builder/id6790187323) All apps by Dodecki Labs: https://appshunter.io/developer/1571436485 ## Related topics [hodapp](https://appshunter.io/ios/topics/hodapp) · [llmconnect](https://appshunter.io/ios/topics/llmconnect) --- *Data collected daily from the US App Store and indexed by [AppsHunter](https://appshunter.io/). User reviews are verbatim App Store reviews. Ratings, prices and chart positions refresh continuously; this snapshot is from 2026-09-04.*