Skip to content

Capri 1.1.0

Choose a tag to compare

View all tags
@Overyx Overyx released this 14 Nov 15:40
v1.1.0
dae1043
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Capri 1.1.0 introduces CapriCore and Skills, a new local inference backend, and improvements to Setup and daily use.

  • CapriCore and Skills. Enable and configure each skill independently. Keep the shortcuts you use and send only relevant skill context to your model.
  • Local and cloud models. Local inference now uses llama.cpp with CPU or Vulkan. Pick separate devices for the language and voice models, use adaptive load controls, or connect to an API. Native NPU support is experimental.
  • Voice and avatar. Added optional external TTS, experimental Russian Qwen3-TTS with LoRA, and voice and microphone controls. Headpats keep the current pose, and the avatar uses the 1.0.4 eye artwork on the updated poses.
  • Desktop and productivity. See verified task checkpoints during computer control. Screen Observation captures changes and preserves more image detail. Added editor integrations, web search controls, editable memories and task recording improvements.
  • Installer and updates. Setup can reuse installed model files, shows per-file download progress, warns clearly about insufficient disk space and resumes large voice downloads. Windows 10 1809–21H2 support is experimental. Windows 10 22H2 and Windows 11 are recommended.

Installation: download and run CapriSetup-1.1.0.exe
Updating from 1.0.4 or an experimental build: Close Capri, run CapriSetup-1.1.0.exe and choose your existing Capri folder to keep your settings, history and compatible models. Setup may download the new local runtime and selected components. The Windows installer is unsigned. SHA-256 checksums are attached.

Full changelog since 1.0.4 · Source changes

Assets 8
Loading