Model access & inference

Advania ALT exposes an OpenAI-compatible inference API at https://api.alt.advania.ai/v1. Point the OpenAI SDK (or any OpenAI-compatible client) at that base URL, use your API key, and you can call the models your account is authorised for.

In this section

• Chat completions
• Streaming responses
• Tools and function calling
• Structured output
• Embeddings
• Model catalog

Every request flows through platform Guardrails before it reaches a model.

Full request/response shapes for each endpoint are in the interactive OpenAPI.