Build powerful AI applications with our search API
AnySearch provides a unified API for aggregating search results from multiple premium data sources. Our intelligent routing system routes queries to the most relevant sources, reducing token waste by 60%+.
Base URL
https://api.agentsearch.tech/v1
All API requests require authentication using an API key. You can generate API keys in your dashboard.
Example Request
curl -H "Authorization: Bearer YOUR_API_KEY" \
https://api.agentsearch.tech/v1/search?q=hello
The search endpoint aggregates results from multiple data sources.
Endpoint
POST /v1/search
| Parameter | Type | Description |
|---|---|---|
| query | string | Search query (required) |
| max_results | integer | Max results to return (1-50, default: 10) |
| sources | array | Optional: filter by source IDs |
curl -X POST "http://localhost:12121/api/v1/search" \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"query": "AGI 2026", "max_results": 5}'
All responses are returned in JSON format with standardized structure.
Success Response
{
"code": 0,
"message": "success",
"request_id": "req_abc123",
"data": {
"results": [
{
"title": "Page Title",
"url": "https://example.com/page",
"snippet": "Brief summary of the page content...",
"content": "Full content or extended summary...\n\ndate: Sep 29, 2025"
}
],
"metadata": {
"total_results": 10,
"search_time_ms": 2357
}
}
}
The content field contains the snippet plus a date suffix in format:
\n\ndate: Sep 29, 2025
Extract with regex: r'\ndate:\s*(.+)'
The API may occasionally timeout (5s). For production use, implement retry logic:
Error responses follow the format: {"code": non_zero, "message": "error description"}
| HTTP Code | Error Code | Meaning |
|---|---|---|
| 401 | 401 | Invalid or missing API key |
| 402 | 402 | Quota exceeded |
| 429 | 429 | Rate limit exceeded |
| 500 | 500 | Internal server error |
Rate limits vary by plan:
Add AnySearch to your AI agent system prompt for web search capabilities.
Skill URL
http://localhost:12121/api/v1/search/mcp-skill
## AnySearch API Access
Base URL: http://localhost:12121
Authentication: Bearer Token
Token: sk_test_-bJL13jBUUb5clSUpz2tphncm-DklaxyO-L_TPrwS4k
Example curl command:
curl -X POST "http://localhost:12121/api/v1/search" \
-H "Authorization: Bearer sk_test_-bJL13jBUUb5clSUpz2tphncm-DklaxyO-L_TPrwS4k" \
-H "Content-Type: application/json" \
-d '{"query": "your search query here", "limit": 10}'
API Endpoints:
• POST /api/v1/search - Search with AI routing
• GET /api/v1/search/sources - List available sources
• GET /api/v1/search/mcp-skill - Get skill manifest
For Claude Code agents, use the skill file:
.claude/skills/anysearch.md