API Documentation

FlashCrawl API

Powerful web scraping and data extraction API with AI-powered structured output.

Getting Started

Base URL

All API requests should be made to the following base URL:

API Base URL

https://api.flashi.ai

Complete endpoint URLs are formed by combining the base URL with the endpoint path. For example: https://api.flashi.ai/v2/scrape

Authentication

All API requests require authentication using a Bearer token in the Authorization header.

Request Headers

Authorization: Bearer YOUR_API_KEY

Scrape

Extract clean content from web pages

GET

/v2/scrape

Scrape URL Content

Retrieve cleaned Markdown content from a URL. Returns HTML by default.

Parameters

Name	Type	Required	Description
url	string	Required	The URL of the website to scrape.
format	string	Optional	Output format. Use "json" to receive a JSON object containing the markdown.
scroll	boolean	Optional	Try scrolling to reveal content. Defaults to true. If false, the page will be loaded as is and quicker response time.

Example Request

GET /v2/scrape?url=https://example.com&format=json&should_click_elements=true
Authorization: Bearer YOUR_API_KEY

Extract

Extract structured data using AI prompts

POST

/v2/extract

Extract Structured Data

Scrape content and extract structured data based on a prompt and JSON schema.

Parameters

Name	Type	Required	Description
url	string	Required	The URL to scrape and parse.
prompt	string	Required	Prompt to guide data extraction (e.g., 'Extract product details').
response_format	object	Required	JSON object describing the desired output structure (OpenAI function calling format).
scroll	boolean	Optional	Try scrolling to reveal content. Defaults to true. If false, the page will be loaded as is and quicker response time.
include_markdown	boolean	Optional	Include the cleaned Markdown content in the response. Defaults to false.

Example Request

POST /v2/extract
Content-Type: application/json
Authorization: Bearer YOUR_API_KEY

{
  "url": "https://example-product-page.com",
  "prompt": "Extract the product name, price, and description.",
  "response_format": {
    "type": "object",
    "properties": {
      "product_name": { "type": "string", "description": "Name of the product" },
      "price": { "type": "number", "description": "Price of the product" },
      "description": { "type": "string", "description": "Product description" }
    },
    "required": ["product_name", "price"]
  },
  "should_click_elements": false,
  "include_markdown": true
}

Selectors

Generate a selector plan once with AI, then extract without AI

POST

/v2/scrape/selectors

Generate Selector Plan

Use AI to infer a reusable CSS-selector plan for a page. Costs 10 credits. Run this once per page layout, then reuse the plan with /v2/scrape/selector-extract.

Parameters

Name	Type	Required	Description
url	string	Required	The page to learn selectors from.
fields	string[]	Optional	Field names to generate selectors for. Omit to let the model choose.
prompt	string	Optional	Extra guidance for what to extract.
wait_for_selector	string	Optional	Wait for this selector to appear before reading the DOM.
scroll	boolean	Optional	Scroll to reveal lazy-loaded content.
validate_plan	boolean	Optional	Also run the plan once and return a sample of the extracted data.

Example Request

POST /v2/scrape/selectors
Content-Type: application/json
Authorization: Bearer YOUR_API_KEY

{
  "url": "https://example.com/products/123",
  "fields": ["product_title", "product_price", "product_images"]
}

POST

/v2/scrape/selector-extract

Extract with Selectors

Apply a selector plan to a page to extract data deterministically — no AI runs, so it is fast and cheap. Costs 2 credits. Expression syntax: 'sel', 'sel::text', 'sel::attr(href)', 'sel::html'; append [] to collect all matches as an array.

Parameters

Name	Type	Required	Description
url	string	Required	The page to extract from.
selectors	object	Required	Map of field name to an array of selector expressions (most specific first).
wait_for_selector	string	Optional	Wait for this selector before reading the DOM.
scroll	boolean	Optional	Scroll to reveal lazy-loaded content.

Example Request

POST /v2/scrape/selector-extract
Content-Type: application/json
Authorization: Bearer YOUR_API_KEY

{
  "url": "https://example.com/products/456",
  "selectors": {
    "product_title": ["h1::text"],
    "product_price": [".price::text"]
  }
}

Screenshot

Capture a PNG screenshot of a page in a real browser

GET

/v2/screenshot

Capture Screenshot

Render a page in a real browser and return a PNG screenshot as a base64 data URL. Costs 2 credits.

Parameters

Name	Type	Required	Description
url	string	Required	The URL to screenshot.
width	number	Optional	Viewport width in pixels.
height	number	Optional	Viewport height in pixels.

Example Request

GET /v2/screenshot?url=https://example.com
Authorization: Bearer YOUR_API_KEY

Tasks

Schedule and manage automated scraping tasks

GET

/v2/tasks

Get Tasks

Get all scheduled tasks for the authenticated user.

Example Request

GET /v2/tasks
Authorization: Bearer YOUR_API_KEY

POST

/v2/tasks

Create Task

Create a new scheduled task to run a scraper.

Parameters

Name	Type	Required	Description
scraper_id	string	Required	ID of the scraper to run.
task_name	string	Required	A descriptive name for the task.
cron_minute	string	Required	Cron expression: minute (0-59 or *).
cron_hour	string	Required	Cron expression: hour (0-23 or *).
cron_day_of_month	string	Required	Cron expression: day of month (1-31 or *).
cron_month	string	Required	Cron expression: month (1-12 or *).
cron_day_of_week	string	Required	Cron expression: day of week (0-6 or *, Sunday=0).
cron_timezone	string	Optional	Timezone for the schedule (e.g., 'America/New_York'). Defaults to UTC.

GET

/v2/tasks/{task_id}

Get Task By ID

Get details of a specific task by its ID.

Parameters

Name	Type	Required	Description
task_id	string	Required	The ID of the task to retrieve (path parameter).

DELETE

/v2/tasks/{task_id}

Delete Task

Delete a specific scheduled task.

Parameters

Name	Type	Required	Description
task_id	string	Required	The ID of the task to delete (path parameter).

POST

/v2/tasks/{task_id}/run

Run Task

Manually trigger a run of the specified task.

Parameters

Name	Type	Required	Description
task_id	string	Required	The ID of the task to run (path parameter).

GET

/v2/tasks/{task_id}/runs

Get Task Runs

Get a list of all historical runs for a specific task.

Parameters

Name	Type	Required	Description
task_id	string	Required	The ID of the task (path parameter).

Scrapers

Manage scraper configurations

GET

/v2/scrapers

Get Scrapers

Get all scraper configurations for the authenticated user.

POST

/v2/scrapers

Create Scraper

Create a new scraper configuration.

Parameters

Name	Type	Required	Description
scraper_name	string	Required	A name for this scraper configuration.
schema_id	string	Required	ID of the schema defining the desired output structure.
scraped_url	string	Required	The target URL or URL pattern for the scraper.
prompt	string	Required	The prompt guiding the data extraction process.

GET

/v2/scrapers/{scraper_id}

Get Scraper by ID

Get details of a specific scraper configuration.

Parameters

Name	Type	Required	Description
scraper_id	string	Required	The ID of the scraper to retrieve (path parameter).

POST

/v2/scrapers/{scraper_id}/run

Run Scraper

Manually trigger a run of a specific scraper configuration.

Parameters

Name	Type	Required	Description
scraper_id	string	Required	The ID of the scraper to run (path parameter).

Schemas

Define data extraction schemas

GET

/v2/schemas

Get Schemas

Get all extraction schemas for the authenticated user.

POST

/v2/schemas

Create Schema

Create a new extraction schema.

Parameters

Name	Type	Required	Description
schema_name	string	Required	A descriptive name for the schema.
schema_json	object	Required	JSON object defining the extraction structure (OpenAI function call format).

GET

/v2/schemas/{schema_id}

Get Schema by ID

Get details of a specific extraction schema.

Parameters

Name	Type	Required	Description
schema_id	string	Required	The ID of the schema to retrieve (path parameter).

MCP

Model Context Protocol server for LLM-optimized content extraction

POST

/mcp

Extract Web Content

MCP server tool for extracting clean HTML/markdown content from web pages, optimized for LLM ingestion.

Parameters

Name	Type	Required	Description
url	string	Required	The website URL to scrape and extract content from.
scroll	boolean	Optional	Enable scrolling to reveal dynamically loaded content. Defaults to true.

Example Request

MCP Client Configuration:

{
  "mcpServers": {
    "flashcrawl": {
      "command": "python",
      "args": ["-m", "mcp", "install", "https://api.flashi.ai/mcp"],
      "headers": {
        "Authorization": "Bearer your-api-key-here"
      }
    }
  }
}