Five minutes from zero to a streaming chat completion through the Aetherin API. All routes live under https://core.aetherin.org/v1.
{ "data": [...], "has_more": false }.curl -sS "https://core.aetherin.org/v1/models" \ -H "Authorization: Bearer $AETHERIN_KEY"
dolphin-3-8b. Non-streaming request:curl -sS "https://core.aetherin.org/v1/chat/completions" \
-H "Authorization: Bearer $AETHERIN_KEY" \
-H 'content-type: application/json' \
-d '{
"model": "dolphin-3-8b",
"messages": [{"role":"user","content":"Hello!"}]
}'"stream": true and use curl -N so output is not buffered.curl -sN "https://core.aetherin.org/v1/chat/completions" \
-H "Authorization: Bearer $AETHERIN_KEY" \
-H 'content-type: application/json' \
-d '{
"model": "dolphin-3-8b",
"messages": [{"role":"user","content":"Hello!"}],
"stream": true
}'https://aetherin.app/playground to chat with models interactively before wiring up your own client.402 with code budget_exceeded. Budgets roll over monthly.rpm_limit returns 429 with code rate_limit_exceeded.{
"error": {
"code": "budget_exceeded",
"message": "monthly budget exhausted",
"type": "payment_required",
"param": null,
"request_id": "req_01HXYZ"
}
}Branch on error.code, never on message. Quote request_id in support requests. See Error Codes for the full table.
List endpoints take limit (1–200, default 50) and an opaque cursor. Stop when has_more is false.
curl -sS "https://core.aetherin.org/v1/keys?limit=50" -b cookies.txt
# → { "data": [...], "next_cursor": "key_01HXYZ", "has_more": true }
curl -sS "https://core.aetherin.org/v1/keys?limit=50&cursor=key_01HXYZ" -b cookies.txt