Getting started · Developers · Applies to 5.0.x

Developer overview

If your tooling already speaks OpenAI-style APIs, Eldric is designed to feel familiar: point the client at your Eldric endpoint, send an API key and call models, chat completions or embeddings.

On this page

Start with the compatible surface Minimal request shape Beyond chat

Start with the compatible surface#

GET/v1/models

List the models available to your key and tenant.

POST/v1/chat/completions

Create a chat completion. Streaming is supported when enabled in the request.

POST/v1/embeddings

Create embeddings for retrieval and search workflows.

Minimal request shape#

curl

curl -X POST https://<your-eldric-host>/v1/chat/completions \
  -H "X-API-Key: $ELDRIC_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"default","messages":[{"role":"user","content":"Summarize this policy."}]}'

Beyond chat#

Use the public Eldric API for conversations, knowledge-base search, agent sessions, media transcription, tenant settings and quota visibility. Keep tenant boundaries explicit in every integration and avoid assuming cross-tenant visibility.

Public API reference RAG guide Help and support