Quickstart
Register, connect a provider, and send your first request.
Set up your organization
Create an account
Every account belongs to an organization — the boundary for tenancy, billing, and RBAC in Inferect Labs. You can invite teammates as members once it exists.
Add a provider credential
Inferect never proxies traffic on shared keys. Add at least one credential under Dashboard → BYOK; it's encrypted (AES-256-GCM) before it touches storage.
Generate an API key
This is what authenticates your app's calls to the data plane, distinct from your dashboard session. Create one under Dashboard → API Keys.
Add a credential via the API
curl -X POST https://api.inferect.online/v1/credentials \
-H "Authorization: Bearer <SESSION_TOKEN>" \
-H "Content-Type: application/json" \
-d '{
"provider": "openai",
"purpose": "chat",
"secret": "sk-..."
}'Send your first request
Same shape as the OpenAI Chat Completions API — because that's exactly what it is.
curl -X POST https://api.inferect.online/v1/chat/completions \
-H "Authorization: Bearer <INFERECT_API_KEY>" \
-H "Content-Type: application/json" \
-d '{
"model": "gpt-4o-mini",
"messages": [{ "role": "user", "content": "Say hello in one sentence." }]
}'Or point an existing OpenAI SDK at it — no new client library to learn.
from openai import OpenAI
client = OpenAI(
base_url="https://api.inferect.online/v1",
api_key="<INFERECT_API_KEY>",
)
resp = client.chat.completions.create(
model="gpt-4o-mini",
messages=[{"role": "user", "content": "Say hello in one sentence."}],
)
print(resp.choices[0].message.content)See why it routed there
Call /v1/routing/explain with the same body and no request actually fires — you get the ranked candidates, the strategy used, and the estimated cost/latency/savings behind the pick.
You're live
From here: try the Playground in the dashboard for interactive testing, or read System Architecture to understand what happens between steps 2 and 3 above.