Offline · Privacy-First · No API Fees

iRun: Private Local AI for Mac, iPhone & iPad

The only no API fee AI engine built for Apple Silicon. Chat with offline LLMs, generate images with CoreML Stable Diffusion, and deploy a private AI agent that runs locally by default. Optional cloud providers when you need more.

100% On-Device Works Offline No API Costs Free to Start
Get iRun Free See How Local AI Works

The local AI for Mac and iPhone that runs entirely on your hardware

0
% Private
0
Data Centers
0
Telemetry
0+
AI Capabilities

The Why

Cloud AI Has a Cost. Local AI Fixes It.

Every cloud chatbot depends on datacenters, Wi-Fi, and endless electricity. iRun puts capable AI on the Mac, iPhone, and iPad you already own: private, unlimited, customizable, and far kinder to the planet.

The Cloud AI Problem

  • Massive datacentersEvery prompt travels to remote server farms built to run 24/7, whether you need them or not.
  • Wi-Fi dependencyNo internet, no AI. Cloud tools stop working the moment you go offline, on a plane, or into focus mode.
  • Enormous electricityTraining and serving cloud models consumes power at industrial scale, driving up global energy demand.
  • Heavy water usageDatacenters rely on vast amounts of water for cooling. AI growth is adding real environmental pressure.

The iRun Answer

  • AI on hardware you ownRun models on your Mac, iPhone, or iPad. No new infrastructure. No datacenter in the loop.
  • On-par local intelligenceiRun tunes small, efficient models for Apple Silicon so everyday chat, coding, and creativity feel as capable as cloud tools.
  • Small models, real powerYour device sets the ceiling, but iRun's job is to make compact local models genuinely useful, not just a fallback.
  • Eco-friendly by designLess datacenter demand, less always-on cooling, less waste. Local AI is one of the smartest sustainability choices you can make.
€

No API Costs

Local chat is free with no tier restriction. Zero per-token bills on every plan.

∞

No Usage Limits

No daily caps, no credit packs, no throttling. Your hardware is the only limit.

⚙

Full Control

Pick your models, own your data, and decide exactly what runs on your device.

🎨

Full Customizability

Themes, personas, system prompts, and workflows shaped around you, not a corporate default.

🌿

Eco-Friendly

Skip the datacenter pipeline. Run AI where you already have silicon, not where the grid strains hardest.

☀

Works Offline

Airplane mode, remote cabins, secure labs. iRun keeps working when the cloud cannot.

Features

Local AI Built for People Who Refuse the Cloud

Stop renting intelligence from servers that log your prompts. iRun is the local AI for Mac, iPhone, and iPad that keeps every conversation, file, and image on your device.

Absolute Privacy, Zero Exceptions

Your prompts, documents, and creative work never touch a server. iRun is a secure offline AI alternative with no telemetry, no training on your data, and no account required.

GDPR Friendly Zero Telemetry

Truly Offline AI

Download once, run forever without Wi‑Fi. Run offline LLMs on iPhone, iPad, or Mac in airplane mode. Perfect for travel, focus, and air-gapped security.

One App, Every Apple Device

Mac, iPhone, and iPad in one seamless local AI experience. Your models and history stay on the devices you already trust.

Your Models, Your Rules

Import any model you choose. No walled garden. Pick the intelligence that fits your hardware and workflow, not a corporation's approved list.

Make It Yours

Custom themes, personas, fonts, and system prompts let you shape iRun into the AI assistant you actually want to use every day.

Blazing Speed on Apple Silicon

Optimized for M-series chips and the Apple Neural Engine for responsive, local inference that feels instant without paying per-token API fees.

Apple Silicon Neural Engine No API Fees

Privacy & Security

The Secure Offline AI Alternative You've Been Looking For

Cloud AI providers monetize your prompts through logging, fine-tuning, and analytics. iRun is architecturally incapable of that. There is no server, no pipeline, and no data leaving your device. If you searched for a private AI agent or secure local alternative to ChatGPT, this is it.

No cloud processingEvery model runs in your device memory, not a remote data center
Zero telemetryNo usage analytics, crash reporting, or behavioral tracking
No training on your dataYour chats, images, and code never improve someone else's model
No account requiredDownload from the App Store and start. No email, no login wall
💻
Your Mac
🧠 LLM
🎨 Image Gen
🤖 Agent
💬 Chat
✗ No data leaves
☁️
Cloud / Servers
BLOCKED

How It Works

Get Local AI Running in Three Steps

No DevOps, no API keys, no terminal. iRun delivers blazing-fast offline chat powered by Apple Silicon optimization, so every response is instant and free.

01

Pick Your Intelligence

Browse and download models right inside the app. From lightweight models for iPhone to powerhouse models for Mac, choose what fits your device, not a vendor's pricing tier.

Model Browser
🌿
Bonsai 8B
1-bit · 1.2 GB · Recommended
⚡
Amber Fable 2.0
Q8_K_M · 2.1 GB · Smart & Light
Available
02

Chat Offline, Instantly

Hold long conversations, reason through problems, and write code with blazing-fast responses that never cost a dime in API tokens because everything runs on your hardware.

N
Nova ✦
● On-device · No internet
What is quantum entanglement?
N
Two particles linked no matter the distance. Measure one, you instantly know the other. Einstein called it "spooky action at a distance." ✨
🔒 Local ⚡ 42 tok/s
Give me a real world example
N
Ask anything...
↑
03

Unlock the Full AI Studio

Generate images with on-device Stable Diffusion, deploy a private AI agent with a 120Hz ProMotion Subagent Bar, build visual workflows, and design in 2D/3D, all without the cloud.

🎨
Image Generation
CoreML Stable Diffusion
✓
🤖
Private AI Agent
Autonomous coding & tools
✓
🔬
DeepSearch
Web research + citations
✓
🔀
Workflows
Visual node-based pipelines
✓

Capabilities

One App. Every Local AI Tool You Need.

From offline chat to autonomous agents and CoreML Stable Diffusion image generation, iRun replaces a stack of cloud subscriptions with one private engine.

Offline Chat That Never Bills You

Blazing-fast, on-device conversations with no API fees and no internet required. Reason through problems, write code, or brainstorm. Your local AI for Mac and iPhone remembers every turn.

  • ✦ Long multi-turn memory for deep work sessions
  • ✦ Rich markdown & code rendering
  • ✦ Attach images for vision-powered chat
  • ✦ Custom AI names, personas & avatars
  • ✦ Searchable full chat history
N
Nova ✦
● Bonsai-8B · 0 internet
Bonsai 8B
Write me a haiku about local AI
N
No cloud above me,
Thoughts bloom in silicon veins,
Yours, only yours. ✨
🔒 Local ⚡ 44 tok/s
Now explain transformer architecture
N
🔒 Local ⚡ 0 tok/s
Ask anything...
↑

CoreML Stable Diffusion, On Your Device

The CoreML Stable Diffusion app experience you wanted: generate unlimited images on Mac or iPhone with no Midjourney subscription, no credit caps, and no uploads to the cloud.

  • ✦ SDXL & SD models accelerated by Apple Neural Engine
  • ✦ 100% private. Images never leave your device
  • ✦ Fine-tune samplers & guidance scale
  • ✦ Built-in gallery with session history
  • ✦ Export to Photos or Finder instantly
Image Generation Denoising
"A futuristic city at golden hour, cinematic, 8K"
1
Denoising step 1 of 10
✓ Saved to gallery
SDXL 10 steps CFG 7.5 🔒 Local

Your Private AI Agent

Hand off a task and watch iRun plan, code, and iterate autonomously. A private AI agent that reads your files, runs commands, and fixes bugs without sending a byte to the cloud.

  • ✦ Autonomous plan-and-execute loops
  • ✦ Read, write & edit project files
  • ✦ Run shell & Python locally
  • ✦ Git-aware code search & diffs
  • ✦ 120Hz ProMotion Subagent Bar
A
Agent ✦
● ReAct mode · iteration 3/10
Fix the auth bug in my Swift project
A
✓ Reading AuthManager.swift
✓ Found race condition at L47
◌ Applying fix...
○ Running tests
○ Verifying build
AuthManager.swift
- await session.validate()
+ await session.validateSync()
$ swift test --filter AuthTests |
🔒 Local Step 3 of 10
Subagents
Planner Coder Tester
120Hz
Give the agent a task...
↑

IdeaGen: From Spark to Blueprint

Turn a rough concept into a structured design document in seconds. Answer a few prompts and get a ready-to-build brief you can hand straight to your agent or image studio.

  • ✦ App, Story, Game, Design & more
  • ✦ Generates structured design docs
  • ✦ Hands off directly to Agent or Image Gen
  • ✦ Complexity & scope sliders
  • ✦ Saves to history automatically
IdeaGen
App
Story
Game
Music
Design
Your interests
Productivity, minimalism, AI
Complexity
Medium
Generated Idea
Send to Agent
Send to Image

DeepSearch: Research on Autopilot

Ask a question and get a cited report back. iRun searches the web, reads sources, and synthesizes findings like a research assistant, with local privacy for everything it processes on-device.

  • ✦ Multi-step autonomous web research
  • ✦ Live web search integration
  • ✦ Source citations & references
  • ✦ Full markdown report output
  • ✦ macOS & iPhone
DeepSearch 🌐 Web
"Impact of local AI on data privacy regulations"
○Searching web: "local AI privacy"
○Reading 8 sources...
○Synthesizing findings...
○Generating report
arxiv.org techcrunch.com wired.com +5 more
Executive Summary
[1] arxiv.org · [2] techcrunch.com

2D & 3D Design Studio

Describe a UI or 3D scene in plain English and watch it render live. Export PNG, OBJ, or USDZ for AR. Design and AI side by side, entirely offline.

  • ✦ AI-driven 2D canvas generation
  • ✦ 3D SceneKit model generation
  • ✦ Export PNG, OBJ, USDZ (AR Ready)
  • ✦ Import existing models
  • ✦ Chat + canvas side-by-side layout
Design Studio
2D3D
3D Scene · SceneKit
Make a metallic sphere with an orbit ring
✦ Generated. Drag to rotate.

GraphGen: Data to Charts in Seconds

Describe your data in plain language and get polished charts instantly. Bar, line, pie, and scatter plots rendered locally with no spreadsheet uploads to the cloud.

  • ✦ Bar, Line, Pie & Scatter charts
  • ✦ Natural language data input
  • ✦ Rendered live in-app
  • ✦ Themed to match your app theme
  • ✦ Chat-driven iteration
GraphGen
Weekly Productivity · Local AI
Describe your data...
↑

Story Builder: Write With AI, Keep It Private

Collaborate on novels, scripts, and poetry chapter by chapter. Set your style and creativity level. Every word stays saved locally on your device.

  • ✦ Page-by-page story creation
  • ✦ Writing styles: Novel, Script, Poetry & more
  • ✦ Creativity level slider
  • ✦ Full story history & session manager
  • ✦ Export & share your work
Story Builder Page 2 of 5
Write
Outline
Pages
Page 2 of 5
Chapter 2: The Signal
The terminal blinked once. Mira leaned closer, certain she had imagined it. But there it was: a pattern in the static, repeating...
Prev
Continue
Next

Customization

An AI Assistant That Feels Like Yours

Themes, personas, fonts, and model controls let you shape iRun into the offline AI experience you actually want to open every day.

N
Nova ✦
● On-device · No internet
Hey, write me a haiku!
N
No cloud above me,
Thoughts bloom in silicon veins,
Yours, only yours. ✨
Nova ✦ ⚡ 41 tok/s · 🔒 Local
Now explain gravity simply.
N
Ask anything...
↑
🎨 Theme Space
💬 Bubble Style
✏ Names & Font
🧠 AI Persona
🎚 Model Controls
Temperature0.70
Higher = more creative and random
Top P0.90
Controls token selection diversity
Context Length4,096
Max tokens held in conversation memory
Repeat Penalty1.10
Prevents repetitive outputs

Pricing

No API Fees. Ever.

Local chat is free forever with no tier restriction. Every Pro plan unlocks the full studio. Pick the price point that fits you.

Free
€0forever

Unlimited local chat on every plan. No Pro subscription required for on-device conversations.

  • ✓ Unlimited offline AI chat
  • ✓ On-device model manager
  • ✓ Zero per-token API costs
  • ✓ Bring your own models
  • ✓ Full privacy & offline mode
  • — Image Generation
  • — Agentic Coding
  • — DeepSearch
  • — Workflows
Download Free
Pro Monthly
€5/ month

Full Pro access. Pay as you go.

  • ✦ Everything in Free
  • ✦ Image Generation (GGUF + SD/SDXL)
  • ✦ StoryGen illustrated stories
  • ✦ Agentic Coding (Beta)
  • ✦ Tool usage & file I/O
  • ✦ Workflows (Coming Soon)
  • ✦ DeepSearch
  • ✦ Agent Manager & personas
  • ✦ VideoGen (Coming Soon)
Get Pro Monthly
Pro Quarterly
€13/ 3 months
Save 13% vs monthly

Full Pro access. Billed every 3 months.

  • ✦ Everything in Free
  • ✦ Image Generation (GGUF + SD/SDXL)
  • ✦ StoryGen illustrated stories
  • ✦ Agentic Coding (Beta)
  • ✦ Tool usage & file I/O
  • ✦ Workflows (Coming Soon)
  • ✦ DeepSearch
  • ✦ Agent Manager & personas
  • ✦ VideoGen (Coming Soon)
Get Pro Quarterly
Most Popular
Pro Yearly
€40/ year
Save 33% vs monthly

Full Pro access. Best annual value.

  • ✦ Everything in Free
  • ✦ Image Generation (GGUF + SD/SDXL)
  • ✦ StoryGen illustrated stories
  • ✦ Agentic Coding (Beta)
  • ✦ Tool usage & file I/O
  • ✦ Workflows (Coming Soon)
  • ✦ DeepSearch
  • ✦ Agent Manager & personas
  • ✦ VideoGen (Coming Soon)
Get Pro Yearly
Lifetime
€60one-time
Best Long-Term Value

Full Pro forever. Pay once, no renewals.

  • ✦ Everything in Pro
  • ✦ All future Pro features
  • ✦ Priority support
  • ✦ No recurring charges
Get Lifetime

All Pro plans include the same features. Subscriptions are managed through the App Store. No hidden API charges, ever.

FAQ

Frequently Asked Questions

Everything you need to know about iRun, the private local AI engine for Apple Silicon.

What is iRun?

iRun is a privacy-first AI engine for Mac, iPhone, and iPad. Its core experience runs large language models, CoreML Stable Diffusion image generation, and autonomous agents locally on your device, offline, with no API fees. You can also optionally connect cloud providers when you need them.

Does iRun support cloud AI providers?

Yes. iRun is local-first, but you can connect optional cloud providers including Claude, Gemini, and OpenRouter. Each offers free and paid tiers with their own pricing. Local models remain the default for privacy, offline use, unlimited chat, and zero per-token costs.

Can I run offline LLMs on iPhone?

Yes. Download a model once inside iRun and chat in airplane mode. iRun is built to run offline LLMs on iPhone, iPad, and Mac with no cloud connection required for local models.

Why is local AI better for the environment?

Cloud AI depends on massive datacenters, constant Wi-Fi, and enormous water and electricity use for training and inference. iRun runs optimized models on the Apple devices you already own, reducing datacenter demand while still delivering capable, everyday AI.

Is iRun a good local AI for Mac?

iRun is optimized for Apple Silicon with fast on-device inference, a 120Hz ProMotion Subagent Bar, visual workflow builders, and CoreML image generation. It is designed to make compact local models feel as useful as cloud tools, making it one of the most complete local AI for Mac apps available.

Does iRun charge API fees or per-token costs?

Not for local models. iRun is a no API fee AI app when you run on-device inference. Core offline chat is free forever. If you choose optional cloud providers like Claude, Gemini, or OpenRouter, those services may have their own free or paid API pricing.

How is iRun different from cloud AI?

ChatGPT, Claude, and Gemini are cloud-first: your prompts go to remote servers, require Wi-Fi, and typically bill per usage. iRun's default mode runs locally with unlimited free chat (no tier restriction), full customizability, and no per-message cost. You can still connect cloud providers inside iRun when you want, but you are never forced to.

What devices does iRun support?

iRun runs on macOS 14+, iOS 17+, and iPadOS 17+. Apple Silicon Macs deliver the best performance, but iRun brings capable local AI to every Apple device you own.

Where can I find documentation and support?

Visit our Documentation for setup guides, model recommendations, and subscription help. For bugs or technical issues, contact irunapp@proton.me or see Support.

iRun mascot, your private local AI companion for Apple devices

Download iRun: Free on the App Store

Join thousands who've switched to private, offline AI. Get the local AI for Mac and iPhone with no per-message cost. Local-first, with optional cloud when you want it.

Download for
macOS
Download on the
App Store

macOS 14+ · iOS 17+ · iPadOS 17+ · Apple Silicon optimized