Overview
Scrapewise is a hosted web-scraping platform. You build scrapers that extract structured data from any website, run them on demand or on a schedule, and consume the results through three customer-facing surfaces:
The three surfaces
1. Customer portal — portal.scrapewise.ai
A web UI for building and operating scrapers visually. Use it to:
- Sign in / create an account
- Build a scraper from a URL — point Scrapewise at a page and pick the fields you want
- Define a schema (the shape of the data the scraper produces)
- Run scrapers on demand and watch the job stream
- Browse scraped rows and export to Excel
- Mint API keys for the other two surfaces
If you’ve never used Scrapewise, start here: Sign up + first key.
2. REST API — https://portal.scrapewise.ai/api/scraper-api/...
A standard REST surface for code. Same scrapers, same data, available to any program that can make HTTPS calls. Authenticate with an API key (see API keys) or a Firebase JWT (portal-managed; you get this automatically when signed in to the portal).
Most-used endpoints:
| Endpoint | What it does |
|---|---|
POST /api/scraper-simple/preview-from-url | One-shot preview: scrape a URL and return the fields immediately. No job, no scraper saved. |
GET /api/scraper/list | List your scrapers |
POST /api/scraper | Create a scraper |
GET /api/scraper/{id}/run | Trigger a scrape job |
GET /api/scraper/{id}/stream-job-status | SSE stream — watch a job complete in real time |
GET /api/scraper/data | Paginated read of scraped rows |
For a 5-minute curl example, see REST quickstart.
3. MCP gateway — mcp.scrapewise.ai
A Model Context Protocol endpoint that exposes Scrapewise as a set of tools that AI agents can call directly. Works with Claude Desktop, Claude Code, claude.ai (custom connectors), and any other MCP-compatible client.
Each REST endpoint is surfaced as a typed tool with descriptions, input/output schemas, and annotations (readOnly / destructive). The agent calls them; you don’t write any glue code.
For a 5-minute setup, see MCP quickstart.
Which surface should I use?
| Use case | Surface |
|---|---|
| I want to build my first scraper | Portal |
| I’m integrating Scrapewise into my code | REST API |
| I want Claude (or another LLM) to use Scrapewise | MCP gateway |
| I want to read scraped data into a notebook / BI tool | REST API |
| I’m exploring what fields are scrapable from a page | Portal’s “preview from URL” |
You don’t have to pick one — all three share the same underlying scrapers, schemas, and data.
Next steps
- New to Scrapewise? → Sign up + first key
- Want a curl example? → REST quickstart
- Want to connect Claude? → MCP quickstart
- Curious about the data model? → Concepts