airunner

Why we included this project

AI Runner is a desktop app that pulls together jobs you would normally run as separate tools: generating images, chatting with an AI companion, and speech-to-text and text-to-speech, all offline on your own hardware with no API keys or subscriptions. Its two main pieces are a layered image canvas built on SDXL and Z-Image Turbo, with image-to-image, inpainting, and background removal, and a named, voiced chatbot with a personality and shifting mood, plus long-term memory built from your conversations. Voice input and output are built in, so you can talk to the companion as easily as typing. That suits hobbyists and small teams who want to keep their data local and pick their own models instead of leaning on a cloud service. The layered canvas is practical for iterating on an image in place, and the app doubles as a reference for how TTS, STT, LLM, and image models can be wired together in a single PySide6 project.