Press ⌥C on any text or image to run local AI actions on macOS
By Tanmay Verma, Founder · Last verified 03 Jul 2026
In short
Cai — Press ⌥C on any text or image to run local AI actions on macOS. Best for macOS power users who want a local, keyboard-driven AI assistant, Developers needing quick issue creation from code errors or screenshots, Privacy-conscious users avoiding cloud AI services. Free to use.
See what real users actually say. We scan live discussions, reviews and complaints across the web and hand you an honest verdict — in under a minute.
3 free scans · no card needed · downloadable report
Cai turns the ⌥C key into a universal action trigger for macOS. It's faster than switching to a chatbot and more private than cloud alternatives — but it's macOS-only and rewards users who invest in setup. If you live in the macOS ecosystem and want a local-first, extensible AI action hub, Cai is a standout. For Windows or Linux users, skip it.
Skip Cai if Skip Cai if you need a cross-platform or mobile AI assistant, or if you prefer a turnkey cloud AI chatbot without local model setup.
Compare with: Cai vs Kagi, Cai vs Everlaw, Cai vs Twistly
Last verified: July 2026
How likely is Cai to still be operational in 12 months? Based on 4 signals — momentum (how recently it shipped), wrapper dependency, revenue model, and web presence.
Last calculated: July 2026
How we score →Cai is a free, open-source macOS menu bar app that acts as a local AI action layer. Select any text or image, press the shortcut, and run custom actions: AI prompts, terminal scripts, URL shortcuts, or create GitHub issues and Linear tickets — all without switching apps. It ships with Ministral 3B (MLX) for zero-setup local AI, supports any HuggingFace MLX model, and can connect to Ollama, LM Studio, Apple Intelligence, or OpenRouter. Cai reads your selection contextually and surfaces relevant actions — e.g., selecting a meeting invite offers to create a calendar event, selecting an address opens Maps. You can chain actions into pipelines: translate then replace, summarize then send to Slack. Everything runs locally by default; cloud is optional. Unlike clipboard managers or AI chat apps, Cai acts on your selection inline without forcing app switching. It's keyboard-first, MIT licensed, and has no telemetry or account requirement. The app integrates with GitHub, Linear, Slack, Notion, Apple Shortcuts, and more via custom destinations and community extensions.
Cai is one of the most thoughtful local AI tools on macOS. Its core idea — act on your selection inline — is executed with precision. The built-in Ministral 3B model is good enough for quick summaries and translations, and you can swap in Qwen 2.5 7B or any MLX model for heavier tasks. Action chaining is a standout: you can pipe a selection through an AI prompt, a shell script, and a Slack message in one keystroke. The context-aware actions (e.g., 'Create Calendar Event' on a meeting invite) feel magical. Privacy is uncompromising: no telemetry, no account, local by default. However, Cai is macOS-only, so Windows and Linux users are out of luck. The built-in model is weaker than cloud LLMs for complex reasoning, and while advanced users can connect to more powerful models, that adds complexity. Setup is easy for basic use, but creating custom actions or chains requires reading the docs. The app is open source and free, which is rare among AI tools. For developers and power users who want a private, keyboard-driven AI assistant that lives in the menu bar, Cai is a compelling choice.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Concrete scenarios for the personas Cai actually fits — and what changes day-one when you adopt it.
You see an error message in your terminal or VS Code. You select the error text, press ⌥C, and choose 'Create GitHub Issue'. The AI summarizes the context and fills the issue title and description. You add labels and submit without leaving your editor.
Outcome: Bug report created in seconds with full context, no app switching.
You draft a blog post in a text editor. You select a sentence and press ⌥C, then choose 'Fix Grammar' (runs locally). Then you select another paragraph and run 'Translate to Spanish' using the built-in model.
Outcome: Writing polished and translated inline, all data stays on your Mac.
You often share summarized articles to your team's Slack. You set up an action chain: select a URL, run 'Fetch URL', then 'Summarize', then 'Send to Slack'. You save it as a custom action. Next time you just select the URL and press ⌥C, then pick the chain.
Outcome: Multi-step workflow runs in one keystroke, saving minutes per article.
as of 2026-07-03
as of 2026-07-03
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
For each published Cai tier: who it actually fits, and what it adds vs. the previous tier. Cross-reference the cost calculator above for projected annual outlay.
Free & Open Source
$0
Ideal for
Individual macOS users who want a free, private, local AI assistant with no subscription
What this tier adds
Starting tier: completely free with built-in local AI, no account needed, all features included.
The company stage and team size where Cai's pricing actually pencils out — and where peers do it cheaper.
Cai is completely free and open source — no subscriptions or hidden fees. This makes it ideal for individual users and developers on a budget, especially compared to paid AI assistants like ChatGPT Plus ($20/mo) or Raycast Pro ($10/mo). However, if you need a managed cloud AI with no local setup, Cai isn't it.
How long it actually takes to get something useful out of Cai — broken out by persona, not the marketing-page minute.
For basic use (built-in AI): download, open, and start using ⌥C immediately — under 2 minutes. For custom actions or chains: 10–30 minutes to read the docs and set up. For connecting to Ollama/LM Studio: additional 15–30 minutes if you already have those installed.
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
Full product docs from getcai.app
Get up and running fast from getcai.app
Full product docs from getcai.app
Full product docs from getcai.app
Full product docs from getcai.app
Full product docs from getcai.app
Full product docs from getcai.app
Common stack mates teams adopt alongside Cai, with the specific reason each pairing earns its keep.
Used Cai? Help shape our editorial sentiment research.