nexusflow
Online

API Reference

nexusflow access is organized into three compatible protocols plus one async task API. OpenAI Chat, Anthropic Messages, and Responses API requests all flow through the same model routing, billing, and monitoring pipeline.

Three-Protocol Compatibility
OpenAI Chat, Anthropic Messages, and Responses API share one set of keys, billing, and monitoring.
Streaming Responses
Chat endpoints support SSE streaming, ideal for real-time interaction and low-latency experiences.
Async Tasks
Image and video use a unified task API, ideal for high-latency and high-concurrency scenarios.
Pre-Launch Validation
Combined with rate limits, error codes, and the monitoring page to help you spot traffic risks before production.

Compatible Protocols

Public Gateway Protocol Scope
AvailableOpenAI Chat / Anthropic Messages / Responses APIText, reasoning, vision, coding, and specialized models are called per supported_protocols.
AvailableOpenAI Embeddings / Image Generations / NexusFlow TasksEmbedding, image, and video models use the corresponding endpoints by capability.

This page lists only the public APIs you can currently call directly at https://nexusflow.hk.

Basics

Base URL
https://nexusflow.hk
Authentication
Authorization: Bearer sk-air-xxx
Request Format
Content-Type: application/json
Response Format
JSON / Server-Sent Events / Task Status

Recommended Integration

Chat
Use `/v1/chat/completions` by default, compatible with the OpenAI SDK.
Protocol
If you already have an Anthropic client or need the Responses API format, use the corresponding compatible endpoint.
Tasks
Image and video models prefer `/v1/tasks` and `/v1/tasks/:id`; images also support OpenAI-style `/v1/images/generations`.

API Endpoints

POST/v1/chat/completionsChat Completions endpoint with multi-turn conversations, streaming, and function calling
POST/v1/messagesAnthropic Messages compatible endpoint, ideal for reusing the Anthropic SDK
POST/v1/responsesResponses API with built-in tools (web search, code interpreter, etc.) and multi-turn context management
GET/v1/modelsRetrieve the list of available models and their information
POST/v1/embeddingsEmbedding endpoint that converts text into vector representations
POST/v1/tasksAsync task endpoint (unified submission for image / video)
GET/v1/tasks/:idQuery async task status