Skip to main content

​
REST Endpoints

The OpenClaw Gateway provides HTTP REST endpoints for specific use cases including OpenAI compatibility, webhooks, and tool invocation.

​
OpenAI-Compatible Chat Completions

​
POST /v1/chat/completions

OpenAI-compatible chat completions endpoint. Authentication: 
Authorization: Bearer your-token-here
Request Body:
​
model
string
required
Model identifier (e.g., "openclaw", "gpt-4"). Maps to OpenClaw agent.
​
messages
array
required
Array of message objects with role and content
​
stream
boolean
default:false
Enable streaming responses (SSE)
​
user
string
User identifier for session scoping
Response:
​
id
string
Completion ID (format: chatcmpl_<uuid>)
​
object
string
"chat.completion" or "chat.completion.chunk" (for streaming)
​
created
number
Unix timestamp
​
model
string
Model identifier from request
​
choices
array
Array of completion choices
​
usage
object
Token usage statistics
Example (Non-Streaming): 
curl -X POST http://localhost:18789/v1/chat/completions \
  -H "Authorization: Bearer your-token" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "openclaw",
    "messages": [
      {"role": "system", "content": "You are a helpful assistant."},
      {"role": "user", "content": "What is OpenClaw?"}
    ]
  }'
Response: 
{
  "id": "chatcmpl_123e4567-e89b-12d3-a456-426614174000",
  "object": "chat.completion",
  "created": 1234567890,
  "model": "openclaw",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "OpenClaw is an open-source AI agent framework..."
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 0,
    "completion_tokens": 0,
    "total_tokens": 0
  }
}
Example (Streaming): 
curl -X POST http://localhost:18789/v1/chat/completions \
  -H "Authorization: Bearer your-token" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "openclaw",
    "messages": [{"role": "user", "content": "Hello"}],
    "stream": true
  }'
Streaming Response (SSE): 
data: {"id":"chatcmpl_123","object":"chat.completion.chunk","created":1234567890,"model":"openclaw","choices":[{"index":0,"delta":{"role":"assistant"}}]}

data: {"id":"chatcmpl_123","object":"chat.completion.chunk","created":1234567890,"model":"openclaw","choices":[{"index":0,"delta":{"content":"Hello"}}]}

data: [DONE]

​
Webhook Endpoints (Hooks)

​
POST /hooks/:path

Custom webhook endpoints for external integrations. Configuration: 
hooks:
  enabled: true
  token: "your-hook-token"
  path: "/hooks"
  maxBodyBytes: 262144
Authentication: 
Authorization: Bearer your-hook-token
Or: 
X-OpenClaw-Token: your-hook-token
Wake Endpoint: 
POST /hooks/wake
Content-Type: application/json

{
  "text": "Wake up message",
  "mode": "now"  // or "next-heartbeat"
}
Agent Endpoint: 
POST /hooks/agent
Content-Type: application/json

{
  "message": "User message",
  "agentId": "default",
  "sessionKey": "hook:session-123",
  "name": "Webhook User",
  "deliver": true,
  "channel": "signal",
  "to": "+1234567890",
  "model": "gpt-4",
  "thinking": "high",
  "wakeMode": "now",
  "timeoutSeconds": 300
}
​
message
string
required
User message to send to agent
​
agentId
string
Target agent ID (defaults to configured default)
​
sessionKey
string
Session key for conversation context
​
name
string
Display name for the sender
​
deliver
boolean
default:true
Whether to deliver response via channel
​
channel
string
Channel to deliver response (e.g., "signal", "telegram", "discord")
​
to
string
Recipient identifier (phone number, user ID, etc.)
​
model
string
Override model for this request
​
thinking
string
Thinking budget: "low", "medium", "high"
​
wakeMode
string
default:"now"
"now" or "next-heartbeat"
​
timeoutSeconds
number
Request timeout in seconds
Response: 
{
  "ok": true,
  "runId": "unique-run-id"
}
Custom Mappings: You can define custom webhook mappings: 
hooks:
  mappings:
    - path: "github"
      action:
        kind: "agent"
        message: "{{ payload.commits[0].message }}"
        channel: "discord"
        to: "#dev-notifications"
See Hooks for details.

​
Tool Invocation

​
POST /api/tools/invoke

HTTP-based tool invocation endpoint. Authentication: 
Authorization: Bearer your-token
Request Body:
​
tool
string
required
Tool name to invoke
​
params
object
Tool-specific parameters
Example: 
curl -X POST http://localhost:18789/api/tools/invoke \
  -H "Authorization: Bearer your-token" \
  -H "Content-Type: application/json" \
  -d '{
    "tool": "weather",
    "params": {
      "location": "San Francisco"
    }
  }'
Response: 
{
  "ok": true,
  "result": {
    "temperature": 68,
    "conditions": "Partly cloudy"
  }
}

​
Channel Endpoints

​
POST /api/slack/events

Slack Events API endpoint for receiving Slack app events. Authentication: Handled via Slack signing secret verification. Request Body: Slack event payload (see Slack Events API). Response: 
{
  "ok": true
}
See Slack Integration for setup.

​
/api/channels/:channelId/*

Channel-specific HTTP endpoints exposed by channel plugins. Authentication: 
Authorization: Bearer your-token
Examples:
  • /api/channels/telegram/webhook - Telegram webhook
  • /api/channels/discord/interactions - Discord interactions
Channel endpoints are dynamically registered by plugins. See individual channel documentation.

​
Health Check

​
GET /health

Health check endpoint (no authentication required). Response: 
{
  "ok": true,
  "status": "healthy"
}

​
Error Responses

All REST endpoints return standard error responses: 400 Bad Request: 
{
  "error": {
    "message": "Invalid request body",
    "type": "invalid_request_error"
  }
}
401 Unauthorized: 
{
  "error": {
    "message": "Unauthorized",
    "type": "authentication_error"
  }
}
429 Too Many Requests: 
{
  "error": {
    "message": "Too many failed authentication attempts. Please try again later.",
    "type": "rate_limited"
  }
}
500 Internal Server Error: 
{
  "error": {
    "message": "Internal error",
    "type": "api_error"
  }
}

​
Next Steps

Hooks

Configure webhook integrations

OpenAI Compatibility

Use OpenAI-compatible API