Integrate autonomous AI into your product. Chat, stream, run multi-day tasks, generate images, and search memory — all via REST API.
Get up and running in under 2 minutes.
npm install -g wispy-aiwispy api create --name "My App" --scopes "chat,marathon"wispy gatewayPass your API key via the Authorization header orX-Api-Key header.
Authorization: Bearer wsk_your_api_key_here # or X-Api-Key: wsk_your_api_key_here
Copy-paste examples in your language of choice.
const response = await fetch(
"https://your-wispy.example.com/api/v1/chat",
{
method: "POST",
headers: {
"Authorization": "Bearer wsk_your_api_key",
"Content-Type": "application/json",
},
body: JSON.stringify({
message: "Summarize the latest AI trends",
session: "my-session-1",
}),
}
);
const data = await response.json();
console.log(data.message.content);All endpoints are prefixed with /api/v1
| Method | Endpoint | Scope | Description |
|---|---|---|---|
| POST | /api/v1/chat | chat | Send a message and get a response |
| POST | /api/v1/chat/stream | chat:stream | Stream a response via SSE |
| GET | /api/v1/sessions | sessions | List all sessions |
| GET | /api/v1/sessions/:key | sessions | Get session message history |
| DELETE | /api/v1/sessions/:key | sessions | Clear a session |
| POST | /api/v1/memory/search | memory | Search agent memory |
| POST | /api/v1/marathon | marathon | Start an autonomous marathon task |
| GET | /api/v1/marathon | marathon:read | List all marathons |
| GET | /api/v1/marathon/:id | marathon:read | Get marathon status |
| POST | /api/v1/marathon/:id/pause | marathon | Pause a marathon |
| POST | /api/v1/marathon/:id/abort | marathon | Abort a marathon |
| POST | /api/v1/generate/image | generate | Generate an image from text |
| GET | /api/v1/skills | skills | List available skills |
| GET | /api/v1/tools | tools | List available tools |
| GET | /api/v1/usage | chat | View API key usage stats |
| GET | /api/v1/health | - | Health check (no auth) |
All responses follow a consistent JSON structure.
{
"id": "req_m1k2j3h",
"object": "chat.completion",
"created": 1738857600,
"message": {
"role": "assistant",
"content": "Here are the latest AI trends..."
},
"thinking": "Let me analyze...",
"tool_calls": []
}{
"error": {
"code": "rate_limit_exceeded",
"message": "Rate limit exceeded. Retry after 45s.",
"retryAfter": 45
}
}
// Error codes:
// unauthorized - Missing/invalid API key
// forbidden - Key lacks required scope
// bad_request - Invalid request body
// rate_limit_exceeded - Too many requests
// not_found - Resource not found
// internal_error - Server errorControl what each API key can access with fine-grained scopes.
chatSend messageschat:streamSSE streamingsessionsRead/manage sessionsmemorySearch agent memorymarathonStart/manage marathonsmarathon:readRead marathon statusskillsList available skillsgenerateImage generationtoolsList/use toolsadminFull access + webhooks*All scopesRate limit info is returned in response headers.
X-RateLimit-Limit: 60 # Max requests per minute X-RateLimit-Remaining: 58 # Remaining requests X-RateLimit-Reset: 1738857660 # Unix timestamp when limit resets
Default: 60 requests/minute. Configurable per API key via wispy api create --rate-limit 120.
Examples of products powered by Wispy API.
Add autonomous AI to your product. Let users ask questions, generate content, or automate workflows.
Build chatbots with memory, tool use, and multi-day task execution — far beyond simple Q&A.
Deploy Wispy as a team copilot that searches docs, writes code, and manages tasks autonomously.
Automate content creation with Marathon mode — research, write, edit, and publish in one task.
Trigger autonomous coding tasks, deployments, and monitoring via API from your CI/CD pipeline.
Connect your iOS/Android app to Wispy for AI features with streaming and real-time responses.