SGLang
Self-hosted SGLang servers for chat, completions and embeddings.
Route requests to your own SGLang servers through DeepintShield. The provider prefix is sgl.
What works
- Chat Completions, and the Responses API through chat
- Text completions and embeddings
- List models, streaming and tool calling
Limits to know: No image generation, speech, transcription, files or batch. The base URL must be configured.
How to connect
Add your SGLang base URL in Model Hub → AI Providers, create a Virtual Key, then call sgl/<model> through https://app.deepintshield.com/v1.
Read the SGLang provider docs. Requests get the same Virtual Key budgets, guardrails and logs as every other provider on the AI gateway.