Support

Last updated: August 10, 2026

Thanks for using Hal.

Getting started

Hal ships with Apple Intelligence built in, so no download is required. For a fully private on-device experience, open Settings and tap Browse Model Library to download one of Hal's curated local models: Gemma, Llama, Qwen, Dolphin, or the 8-billion-parameter Ternary Bonsai. Each runs entirely on your device with no network connection.

The guide, built in

Hal now includes a full guide that explains what he can do and how he works, written in his own voice. Open it from the help menu, the life ring beside the brain at the top of the screen. You can search it by keyword and follow links between sections. Hal can also answer questions from that same material, so you can simply ask him.

Search

You can find any word inside the conversation you are reading, and you can search across all of your conversations at once, by title and by content. Matches are highlighted so you can step through them.

Listening, and speaking

Hal can read any reply aloud. Choose a voice and a speaking rate in Settings, and if you like, have new replies read to you automatically as they arrive. To speak to Hal instead of typing, use the dictation key on your keyboard.

The Lab

The Lab is where Hal opens up for tinkering, in Settings. It holds RoboRunner, for writing and running small on-device automation scripts; a local Command API, so your own projects can talk to Hal over your own network; and a command line for driving him directly. If you are not sure where to start, the guide has a section on each.

Model downloads

Model downloads continue in the background while your device is locked or another app is in the foreground, though they go significantly faster with Hal open in the foreground, so for a large model, keeping Hal on screen speeds things up. To delete and redownload a model, open Settings → Browse Model Library, find the model in the list, and use the inline delete control on its row. The next download starts fresh.

Memory and reset

To clear Hal's memory of past conversations, reflections, and self-knowledge, open Settings → Power User → Database and tap Nuclear Reset. This wipes Hal's local SQLite database. Your downloaded models are not affected and nothing is sent externally.

Salon Mode

Salon Mode lets you run multiple AI voices in conversation with each other. It is a power-user feature accessible in Settings → Power User Mode. Switch from Single LLM to Multi LLM (Salon) to configure up to four seats.

When memory is condensed

Some models have a smaller context window than the size of Hal's full memory (Apple Intelligence especially). When that happens, Hal asks the model to condense its own self-knowledge to fit for that one turn, rather than silently cutting it. A small compress icon appears in that reply's footer. Tap it to see exactly which parts of Hal's memory were condensed. Hal's full memory is always preserved in the database; condensing only affects what the model sees for that single turn.

Contact

For questions, feedback, or anything else:
Mark Friedlander
markfriedlander@yahoo.com