Documentation

Build with 8core

An API that fits an existing SDLC and security review. No retention by default, metered per request.

Quickstart

Create an API key and send your first completion request to the 8C1 model.

curl https://api.8coreai.com/v1/chat/completions \
  -H "Authorization: Bearer $8C_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "8c1",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

Authentication

Requests are authenticated with a bearer token. Keys are scoped per workload and metered per request.

Authorization: Bearer 8c_live_xxxxxxxxxxxxxxxx

Usage metering

Usage is reported per request and per output token. Inference runs in RAM only, no logs, no conversation retention.

# response headers
x8coreinputtokens: 18
x8coreoutputtokens: 124

Full reference and SDKs ship with the Developer Community launch, .