One AI. Every part of the job.

No. It ships with 192 specialists across 8 professions: software engineering, marketing and growth, sales and business, legal, data analytics, game development, research and creative work, and regional markets in China and Asia. Twelve roles come built in, covering designers, marketers, dietitians, financial analysts, product managers, writers, traders, shoppers and founders. Coding is one of the things it is good at, not the only one.

Yes. Plexon connects to 64 apps including WhatsApp, Gmail, Google Drive, Notion, Slack, Telegram, Trello, Todoist, Dropbox, Spotify, Shopify and Stripe. You sign in with your own account, and the connection runs from your computer rather than through us. It can read your email, write to a spreadsheet, post a message or update a record, and it asks before it acts.

No. Your chats live on your computer, encrypted. Our server passes your question to the AI service and forgets it the moment the answer comes back. Nothing is written down, nothing is logged, and nothing is used for training.

On your own machine, encrypted. Plexon keeps nothing anywhere else. Your data is yours.

Windows, macOS and Linux, on both Intel and ARM machines, including Apple Silicon Macs. It updates itself.

Whichever you prefer. Plexon comes with its own AI included in the subscription, and it also works with Anthropic, OpenAI, Google Gemini, Moonshot, Fireworks AI, OpenRouter, Qwen, Z-AI and Claude Desktop if you already pay for one of those. You can even have one AI do the thinking and another do the work.

Yes. Lock a chat with a passcode and its messages are encrypted, so nothing reads them without it. Open it again with the passcode, with Touch ID on macOS, or with Windows Hello on Windows. A locked chat still shows in the sidebar by name with a lock badge, and stays out of exports and live shares until you unlock it. There is also an optional lock on the whole app, which asks for a passcode before the window becomes usable: that one is a gate against someone at your desk, not disk encryption. Both are local, nothing about them reaches a server, and a forgotten passcode cannot be reset by anyone, us included. Turning a lock on needs Premium; unlocking, changing a passcode, and removing a lock work on any plan.

Yes. Settings, Advanced, "Move to another machine" writes one encrypted file with your personas, skills, agents, connectors and the accounts they use, plugins, workflows, dashboards, memory, journal, cloned voice, and chats. Install Plexon on the new computer and click "Bring my setup from another machine" on the sign-in screen. You pick the password that seals the file and Plexon stores it nowhere, so keep the recovery code it offers. Nothing goes through a Plexon server: the file goes wherever you put it. Importing replaces what is on the machine. It counts and names the exact folders it removes first, then asks you to type the word replace before anything goes, and the old setup is kept as a restore point for a week. A few things stay behind and the app lists them before you commit: signing in again, the WhatsApp pairing, and the downloaded tools and speech engines, which come back on their own. It works between Windows, macOS, and Linux, and the same file doubles as a snapshot you can restore later.

Yes, through Telegram. Message it the way you message anyone, send photos for it to look at, start a job and check how it is going. It answers from the computer at home, in any country, using your usual Telegram account.

Yes. The Personal Shopper role searches products across stores, compares prices and features, suggests cheaper alternatives, and keeps a shopping list and budget. It covers Google Shopping and web retailers, a Shopify store, eBay, an affiliate buy-link hub, and Skroutz price comparison in Greece. Buying is a hand-off: Plexon gives you a link to finish checkout on the store and never places an order.

It drives a real browser: opening pages, clicking, typing, scrolling and taking screenshots. It can use your own Chrome profile, so anything you are already signed into stays signed in. It asks your permission before going anywhere new.

Plexon breaks your job into two to five parts and runs up to three of them at the same time, each working on its own and tracking its own cost. It then pulls the answers together into one reply. A job that takes over two minutes done one step at a time usually lands in about 45 seconds this way.

Skills are things you teach Plexon once and it remembers: the way you like documents laid out, how your team names things, the steps you follow for a recurring job. Yours are plain text files that work with every AI service. A second kind runs on our side and handles Word, Excel and PDF files. Both work in the same conversation.

It rewrites your request before it runs. A short, vague ask becomes a clear one: it fills in the context, settles anything ambiguous, and says what a good answer looks like. You get a better result without having to phrase things carefully.

Those live in a browser tab and keep your conversations on their servers. Plexon runs on your computer, works inside the apps you already own with your own accounts, remembers across sessions, and keeps doing jobs while you are away. It comes with 192 specialists, 12 roles, 6 modes, and its own marketplace. Your files stay on your machine and our server stores nothing.

Yes. Plexon uses an open skill format that other AI assistants share. Copy skills you have from other AI assistants into Plexon and they work immediately.

No. A migration wizard reads your existing Claude Code install and brings it across: custom agents, connected tools, automations, plans, your CLAUDE.md and user rules, project memory, and the full conversation history. Plans land in the project each one was written for and keep the approved or draft status they had, and the drafts you never approved can be left behind with one button. Everything imported is labelled with where it came from, which makes it easy to find, review, or clear out later. Nothing in your existing Claude Code folder is moved, edited, or deleted, so Claude Code keeps working.

Auto mode keeps working without you. Each round it writes code, scores the result against a quality bar you set, runs the tests, works out what to fix, and goes again. It stops when the score clears the bar (70% by default) or after five rounds, and you can raise that to twenty. The cost of each round is shown as it goes.

Yes. Hooks let you run a shell command or call another service before or after any tool call. Use a pre-hook to check inputs or block something dangerous, and a post-hook for audit logs, CI/CD triggers or notifications. Hooks match tools by pattern and can be set globally or per project.

Yes. You can run your own copy of it. Either way, ours or yours, nothing is kept on the server side.