Properties
| Name | Type | Description |
|---|---|---|
| prompt | string | Search query prompt |
| tools | array | Tools to search with: `web` (Web Search) and/or `twitter` (X Search). Retired tools are still accepted so older clients keep working, but none of them returns its own result key. `youtube` (`Youtube S |
| start_date | object | The start date for the search query. Format: YYYY-MM-DDTHH:MM:SSZ (UTC) |
| end_date | object | The end date for the search query. Format: YYYY-MM-DDTHH:MM:SSZ (UTC) |
| date_filter | object | Deprecated relative window. When set (and no start_date/end_date), it is translated to start_date/end_date. Prefer start_date/end_date. |
| streaming | boolean | Whether to stream results |
| result_type | object | The result type to be used for the search |
| mode | object | Speed/quality mode for the search (fast, balanced, deep). fast is web-only: X (Twitter) search supports balanced and deep. |
| system_message | object | The system message to be used for the search |
| scoring_system_message | object | System message for scoring the response |
| count | object | The number of results to return per source. Min 10. Max 200. |
| include_domains | object | Restrict Web Search results to these domains. |
| exclude_domains | object | Drop Web Search results from these domains. |
| highlights | boolean | Include highlights (token-efficient key excerpts from each page) on Web Search results. |
| page_text | boolean | Include the full page text on each Web Search result. Responses get significantly larger. |
JSON Schema
{
"$schema": "https://json-schema.org/draft/2020-12/schema",
"$id": "https://artifacts.apis.io/desearch-ai/json-schema/desearch-ai-scraper-streaming-synapse-request-schema.json",
"title": "ScraperStreamingSynapseRequest",
"x-generated": "2026-10-30",
"x-method": "derived",
"x-generator": "derive-json-schema.py",
"x-source": "openapi/desearch-ai-openapi.json#/components/schemas/ScraperStreamingSynapseRequest",
"properties": {
"prompt": {
"type": "string",
"title": "Prompt",
"description": "Search query prompt"
},
"tools": {
"items": {
"anyOf": [
{
"$ref": "#/$defs/ToolEnum"
},
{
"type": "string"
}
]
},
"type": "array",
"title": "Tools",
"description": "Tools to search with: `web` (Web Search) and/or `twitter` (X Search). Retired tools are still accepted so older clients keep working, but none of them returns its own result key. `youtube` (`Youtube Search`) is dropped from the list. `reddit`, `hackernews`, `wikipedia` and `arxiv` (also `Reddit Search`, `Hacker News Search`, `Wikipedia Search`, `ArXiv Search`) become `web` with reddit.com, news.ycombinator.com, wikipedia.org or arxiv.org added to `include_domains`, which pins the whole Web Search in the request to those domains; their results arrive in `search`. A list left with no supported tool, such as `[\"youtube\"]`, is a 422."
},
"start_date": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "Start Date",
"description": "The start date for the search query. Format: YYYY-MM-DDTHH:MM:SSZ (UTC)"
},
"end_date": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "End Date",
"description": "The end date for the search query. Format: YYYY-MM-DDTHH:MM:SSZ (UTC)"
},
"date_filter": {
"anyOf": [
{
"$ref": "#/$defs/DateFilterEnum"
},
{
"type": "null"
}
],
"description": "Deprecated relative window. When set (and no start_date/end_date), it is translated to start_date/end_date. Prefer start_date/end_date."
},
"streaming": {
"type": "boolean",
"title": "Streaming",
"description": "Whether to stream results",
"default": true
},
"result_type": {
"anyOf": [
{
"$ref": "#/$defs/ResultTypeEnum"
},
{
"type": "null"
}
],
"description": "The result type to be used for the search",
"default": "LINKS_WITH_FINAL_SUMMARY"
},
"mode": {
"anyOf": [
{
"$ref": "#/$defs/SearchModeEnum"
},
{
"type": "null"
}
],
"description": "Speed/quality mode for the search (fast, balanced, deep). fast is web-only: X (Twitter) search supports balanced and deep."
},
"system_message": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "System Message",
"description": "The system message to be used for the search",
"default": ""
},
"scoring_system_message": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "Scoring System Message",
"description": "System message for scoring the response"
},
"count": {
"anyOf": [
{
"type": "integer",
"maximum": 200.0,
"minimum": 10.0
},
{
"type": "null"
}
],
"title": "Count",
"description": "The number of results to return per source. Min 10. Max 200.",
"default": 10
},
"include_domains": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"title": "Include Domains",
"description": "Restrict Web Search results to these domains."
},
"exclude_domains": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"title": "Exclude Domains",
"description": "Drop Web Search results from these domains."
},
"highlights": {
"type": "boolean",
"title": "Highlights",
"description": "Include highlights (token-efficient key excerpts from each page) on Web Search results.",
"default": true
},
"page_text": {
"type": "boolean",
"title": "Page Text",
"description": "Include the full page text on each Web Search result. Responses get significantly larger.",
"default": false
}
},
"type": "object",
"required": [
"prompt",
"tools"
],
"$defs": {
"DateFilterEnum": {
"type": "string",
"enum": [
"PAST_24_HOURS",
"PAST_2_DAYS",
"PAST_WEEK",
"PAST_2_WEEKS",
"PAST_MONTH",
"PAST_2_MONTHS",
"PAST_YEAR",
"PAST_2_YEARS"
],
"title": "DateFilterEnum"
},
"ResultTypeEnum": {
"type": "string",
"enum": [
"ONLY_LINKS",
"LINKS_WITH_FINAL_SUMMARY"
],
"title": "ResultTypeEnum"
},
"SearchModeEnum": {
"type": "string",
"enum": [
"fast",
"balanced",
"deep"
],
"title": "SearchModeEnum"
},
"ToolEnum": {
"type": "string",
"enum": [
"web",
"twitter"
],
"title": "ToolEnum"
}
}
}
Work with this as data
Every JSON Schema here is available over the APIs.io API and to AI agents over MCP.
MCP server
One button, every client — Claude, Cursor, VS Code and the rest.
https://apis.io/mcp
Tools for schemas
4 MCP tools reach this
find_json_schemasBrowse and filter every JSON Schema in the catalog.apis_io_searchSTART HERE — APIs, providers and tags for one query, each with its total.resolveTurn a domain, URL or GitHub org into the provider it belongs to.find_cohortsEvery scored population of providers in the catalog.
Call it yourself
curl for this page
This JSON Schema
curl "https://apis.io/api/v1/json-schemas/desearch-ai-scraper-streaming-synapse-request"
All schemas
curl "https://apis.io/api/v1/json-schemas?limit=25"
Discovery needs no key. Ratings and market analysis are Pro.
Get an API key
Free tier, no form to fill in. Signing in shares your email address with us — we store it to create your key and to recognise you if you sign in with another provider. See our Privacy Policy and Terms.
A second provider on the same verified email joins the account you already have.