v1.1.13

Features

• Long responses now continue automatically when a model reaches its output-token limit.
• Reply-detail settings now work across supported models, including GPT-5-compatible APIs.

Bug Fixes

• Fixed provider thought-signature handling and streaming timeouts.
• Submitting a new session now switches to its tab immediately, avoiding a redirect back to the draft.

Downloads

CLI

macOS

• Apple Silicon ZIP | Intel ZIP

Linux

• x64 tar.gz | ARM64 tar.gz

Windows

• x64 ZIP | ARM64 ZIP

Desktop

macOS

• Apple Silicon DMG | Apple Silicon ZIP

Linux

• x64: DEB | AppImage | RPM
• ARM64: DEB | AppImage | RPM

Windows

• x64: Installer | ARM64: Installer

Full Changelog

Compare changes