Skip to main content
SiYuan is a privacy-first personal knowledge management system with block-level Markdown editing and built-in AI writing and Q&A. Its AI settings accept a custom OpenAI-compatible endpoint, so you can point them to the ScitiX Model Inference API. The steps below reflect SiYuan’s current AI settings. Newer versions may differ slightly. For available models, such as glm-5.2 and DeepSeek-V4-Flash, and pricing, see the model catalog.

​
Configure

Create an API Key on the API Keys page first.

​
1. Add the provider

Open Settings → Artificial Intelligence, add a provider through the Add provider dialog, and fill in the required fields. The dialog notes that the actual endpoint is /chat/completions under this address, so the API Base URL ends with /v1. SiYuan — Add provider dialog pointed at ScitiX

​
2. Add the model

Add a model under the ScitiX provider through the Add model dialog. Set Model name to glm-5.2, or click Fetch available models. Click Test connection. Expect Connection successful, then confirm. SiYuan — Add model with connection test passing

​
Verify

Open the AI chat panel, select glm-5.2 from the model selector at the bottom, and send a message. A normal reply confirms that the connection works.
The assistant may identify itself as “the SiYuan AI assistant” rather than the underlying model. That is SiYuan’s own system prompt, not a routing issue; the model selector still shows the ScitiX model in use.
The AI panel is marked as an early-preview feature in current builds. Back up your data before heavy use.
SiYuan — chatting with a ScitiX model

​
FAQ

  • Model not responding: confirm the API Base URL, API Key, and model name. The model must be available over the OpenAI chat-completions protocol, such as glm-5.2.