API ReferenceAPI 参考

oBeaver exposes an OpenAI-compatible HTTP API. Point any OpenAI SDK or compatible client at your local server.oBeaver 提供 OpenAI 兼容的 HTTP API。将任意 OpenAI SDK 或兼容客户端指向本地服务器即可。

Chat Server (obeaver serve)聊天服务 (obeaver serve)

Embedding Server (obeaver serve-embed)嵌入服务 (obeaver serve-embed)

Dashboard Server (obeaver dashboard)仪表盘服务 (obeaver dashboard)

The web dashboard (chat UI + real-time memory gauges) is only available via obeaver dashboard. It is not served by obeaver serve or obeaver serve-embed.网页仪表盘（聊天界面 + 实时内存仪表）仅通过 obeaver dashboard 提供，obeaver serve 和 obeaver serve-embed 不包含仪表盘。

POST /v1/chat/completionsPOST /v1/chat/completions

Create a chat completion. Supports both streaming (SSE) and non-streaming modes.创建对话补全。支持流式（SSE）和非流式模式。

Request Parameters请求参数

Non-streaming Example非流式示例

curl http://127.0.0.1:18000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Phi-4-mini",
    "messages": [
      {"role": "system", "content": "You are a helpful assistant."},
      {"role": "user", "content": "What is ONNX Runtime?"}
    ],
    "max_tokens": 512,
    "temperature": 0.7
  }'

Streaming Example流式示例

curl http://127.0.0.1:18000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Phi-4-mini",
    "messages": [{"role": "user", "content": "Tell me a joke"}],
    "stream": true
  }'

Tool Calling Example工具调用示例

curl http://127.0.0.1:18000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Phi-4-mini",
    "messages": [{"role": "user", "content": "What is the weather in Tokyo?"}],
    "tools": [{
      "type": "function",
      "function": {
        "name": "get_weather",
        "description": "Get current weather for a city",
        "parameters": {
          "type": "object",
          "properties": {"city": {"type": "string"}},
          "required": ["city"]
        }
      }
    }]
  }'

POST /v1/embeddingsPOST /v1/embeddings

Compute text embeddings. Available on the embedding server (obeaver serve-embed).计算文本嵌入向量。在嵌入服务（obeaver serve-embed）上可用。

Request Body请求体

curl http://127.0.0.1:18001/v1/embeddings \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Qwen3-Embedding-0.6B",
    "input": ["Hello, world!", "Embeddings are useful."]
  }'

GET /healthGET /health

Returns the health status of the server, including loaded model name and engine type.返回服务器健康状态，包含已加载的模型名称和引擎类型。

curl http://127.0.0.1:18000/health

GET /api/system/memoryGET /api/system/memory

Returns CPU, GPU, NPU, and process memory statistics as JSON. This powers the web dashboard's real-time gauges.以 JSON 格式返回 CPU、GPU、NPU 和进程内存统计信息。这是网页仪表盘实时仪表的数据源。

curl http://127.0.0.1:18000/api/system/memory

Model Management模型管理

GET /api/models/availableGET /api/models/available

Lists all cached models available for hot-swapping.列出所有可用于热切换的已缓存模型。

POST /api/models/loadPOST /api/models/load

Hot-swap the active model at runtime without restarting the server.运行时热切换活动模型，无需重启服务器。