Local AI path
Core Disp8ch use does not require a hosted model-provider key. Choose Local AI in onboarding, inspect this PC, run the recommended model command yourself, then test and save the endpoint.
The advisor reads hardware and installed runtime signals to rank practical choices. It does not silently download models, start unknown executables, replace active models, or upload your hardware inventory.
Common runtimes
Ollama is the easiest path for many users. LM Studio, llama.cpp, vLLM, and SGLang can work through OpenAI-compatible base URLs.
- Ollama: http://127.0.0.1:11434
- LM Studio: http://127.0.0.1:1234/v1
- llama.cpp server: http://127.0.0.1:8080/v1
- vLLM: http://127.0.0.1:8000/v1
- SGLang: http://127.0.0.1:30000/v1
What still needs credentials
Live web search, external channels, hosted image generation, and third-party APIs need network access and the credentials you choose to configure. Local inference only covers model calls that can run through the local endpoint.
Next step
This doc mirrors the in-app Docs tab. Install Disp8ch, open onboarding, and use WebChat when you want the app to inspect your current setup.