Skip to content

Sato 1.1 — Multi-Provider AI Support

Choose a tag to compare

View all tags
@vitalune vitalune released this 25 Apr 03:28
· 37 commits to main since this release
v1.1.0
cad8ec8

Sato now talks to more than just Claude. Pick from Anthropic, OpenAI, or Ollama, and switch between models without leaving the menu bar.

What's new

Three AI providers, your choice:

  • Anthropic — Claude Opus 4.7, Claude Sonnet 4.6, Claude Haiku 4.5
  • OpenAI — GPT-5.4, GPT-5.4 Mini, GPT-5.4 Nano
  • Ollama — run models locally on your machine, or use Ollama Cloud

Context Profiles can now override your default model. Set Sato to use Claude Sonnet by default, but route your "Code Review" profile through GPT-5.4 Nano because it's cheaper. Or your "Quick Question" profile through a local Ollama model because it's free.

Markdown now renders properly. No more raw **asterisks** cluttering responses.

A few quality-of-life fixes:

  • Permissions screen no longer gets stuck after granting Screen Recording access
  • Cleaner copy on the permissions panel
  • Stable bundle identifier (this is the last release that'll force a one-time permissions re-grant)

Heads up for v1.0 users

This release changes Sato's bundle identifier under the hood, which means macOS will treat it as a new app the first time you launch it. You'll need to grant Accessibility and Screen Recording permissions one more time. Sorry for the friction — after this, future updates will be smooth.

Assets 3
Loading