01
Base URL
Use `https://aergia-ai.com/v1` with OpenAI-compatible clients.
Docs
Call frontier models with one endpoint
Use your existing OpenAI-compatible SDK or HTTP client. Swap the base URL, set your API key, and choose an Aergia AI model ID.
Quickstart
Chat completions
The examples below use the OpenAI-compatible route. Keep your API key private and create separate keys for development and production.
curl https://aergia-ai.com/v1/chat/completions \
-H "Authorization: Bearer $AERGIA_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "anthropic/claude-opus-4.8",
"stream": true,
"messages": [
{"role": "user", "content": "Explain KV cache in one paragraph"}
]
}'
-H "Authorization: Bearer $AERGIA_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "anthropic/claude-opus-4.8",
"stream": true,
"messages": [
{"role": "user", "content": "Explain KV cache in one paragraph"}
]
}'
02
Model ID
Pass one of the available model IDs from the Models page.
03
Streaming
Set `stream: true` for interactive chat and agent workflows.