AI models¶
Hosted open models served from the platform, billed per token.
Manage models¶
AI models lists the models available to your tenant and a catalog to pull more from. Pulling a model is a background task.
Call a model¶
The API is OpenAI-compatible, so existing SDKs only need a different base URL and token. Use a service account token (or your own session token):
curl https://gateway.cloud.rso.dev/ai/v1/chat/completions \
-H "Authorization: Bearer $RSO_DEV_ACCESS_TOKEN" \
-H "X-Tenant: <tenant-id>" \
-H "Content-Type: application/json" \
-d '{"model": "<model-name>", "messages": [{"role": "user", "content": "Hello"}]}'
| Endpoint | Shape |
|---|---|
POST /ai/v1/chat/completions |
OpenAI chat completions, streaming supported |
POST /ai/api/chat |
Ollama chat |
POST /ai/api/generate |
Ollama generate |
POST /ai/api/embed |
Ollama embeddings |
Requests are rate limited per client address.
Usage¶
Prompt and completion tokens are counted per model and per account; the Billing page shows the month's usage per model.