// builtin tools · 03
Voice, TTS, STT and image
Okami has a turn-based voice bridge and image generation/editing through Codex OAuth. Voice is optional (uv sync --extra voice).
params okami transcribe <audio> STT Transcribes with local Whisper. okami say <text> TTS Generates MP3 with Edge TTS. okami voice bridge Microphone → STT → agent → TTS → speaker. okami image <prompt> image Generates a new image. okami image <prompt> --ref img.png image edit Transforms reference images through /images/edits. generate_image tool Harness tool that writes PNG output inside the workspace.
Image generation uses gpt-image-2 with the Codex subscription token. If missing, run okami login codex.