OpenAI-compatible chat completions with streaming.
SENDFU provides authenticated OpenAI-compatible chat-completions services. The API supports JSON responses, SSE streaming, usage metadata, and structured error responses where applicable.
Public API surface
Model discovery is public. Chat-completions requests require a SENDFU-issued Bearer API key and can use only the selected model IDs exposed by the live catalog.
GET /v1/models
Returns model services currently available through SENDFU. It does not add unverified capability, entitlement, pricing, or availability claims.
GET /v1/status
Returns application configuration and request-counter checks only. It does not prove model inference availability.
POST /v1/chat/completions
Processes an authorized request for an allowlisted model through the SENDFU chat-completions service, including supported streaming requests.
Base URL and authentication
Use the same origin as the website. Your request must include a SENDFU-issued Bearer API key; service-side credentials are never accepted from clients or exposed by this service.
List model services
Inspect the live filtered catalog before selecting an ID. The catalog reports the current SENDFU-operated inference services and model provenance; it is not a guarantee of capacity, pricing, entitlement, specific endpoint support, or future availability.
curl https://sendfuglobal.com/v1/modelsChat-completions request
Send a selected model ID and a valid SENDFU Bearer API key. The service rejects IDs outside the current allowlist before beginning service processing.
curl https://sendfuglobal.com/v1/chat/completions \
-H 'Authorization: Bearer [SENDFU_REVIEWER_KEY]' \
-H 'Content-Type: application/json' \
-d '{
"model": "z-ai/glm-5.1",
"messages": [{"role": "user", "content": "Hello"}]
}'Set "stream": true to request an SSE response. The service returns service success and error payloads rather than fabricating inference results.
Service boundary
This endpoint provides authenticated SENDFU-operated inference. SENDFU operates the public gateway, inference runtime, and deployed model services. Model trademarks, developer identity, and base-model ownership remain with their respective owners unless explicitly stated otherwise.