mirror of
https://github.com/screentinker/screentinker.git
synced 2026-06-15 02:33:15 -06:00
How to run the AI design feature fully local + free: Ollama (OpenAI-compatible LLM) for text/layout and stable-diffusion.cpp (Vulkan) for images, plus the SELF_HOSTED requirement for localhost endpoints, an OpenAI fallback, and GPU troubleshooting (incl. the Blackwell CUDA-fails/Vulkan-works note). Linked from the README integrations section. |
||
|---|---|---|
| .. | ||
| android-troubleshooting.md | ||
| local-ai-setup.md | ||
| multi-tenancy-design.md | ||