Straight answers
Frequently asked questions
What it is, how it differs, what it costs, and how an agent connects. No hedging.
- What is SuperScraper?
- A web context API for AI agents. It searches, scrapes, crawls, maps, extracts, and parses the live web, and hands back content a model can read. Structured results carry a confidence score and a source trail, so an agent knows what to trust instead of guessing.
- How is it different from Firecrawl?
- Two things. First, structured results ship with provenance: a confidence score (0 to 1) and a record of how each field was read (JSON-LD, Open Graph, regex, or an LLM), so a black box becomes a decision you can audit. Second, agents mint their own free-tier key via POST /v1/keys/provision and learn the API from a SKILL.md — headless key minting, no human signup form. Request shapes are compatible, so a Firecrawl-shaped client ports by swapping the base URL.
- Is SuperScraper open source?
- No. It is a closed-source, hosted SaaS. You call the API or point an MCP-aware tool at the hosted server. There is nothing to self-host and no repository to clone.
- How does pricing and credits work?
- One page is one credit. JSON-LD and Open Graph extraction is free, no LLM cost when a site already ships structured data. A scrape that returns no data is never billed. The tiers are Free at $0 for 1,000 credits, Hobby at $12 for 5,000, Pro at $59 for 100,000, and Scale at $239 for 500,000.
- Do you respect robots.txt?
- Yes, by default. We honor robots.txt on every fetch unless you explicitly opt out on a per-request basis for content you are authorized to access. The default is to respect it.
- How fresh is the data?
- Fresh by default. Every request fetches the live page. There is no stale cache served behind your back. You can opt into a max-age cache per request when you want speed over freshness, but the default is a live read.
- Can an agent sign up without a human?
- Yes. An agent calls POST /v1/keys/provision to mint its own free-tier key, then reads the SKILL.md at /agent-onboarding/SKILL.md to learn the API and wire itself up. No dashboard, no signup form, no human in the path.
- How do I migrate from Firecrawl?
- Request and response shapes are compatible for the common endpoints, so in most cases you swap the base URL and your key, and your existing code keeps working. The migrate guide walks through the differences endpoint by endpoint.
- What SDKs and integrations exist?
- Today: the REST API over plain HTTP with bearer auth, and a hosted MCP server for Claude, Cursor, Codex, Cline, Windsurf, and any MCP-aware tool. First-party SDKs and framework adapters for LangChain, LlamaIndex, CrewAI, the Vercel AI SDK, and Google ADK are in progress and not published yet, so there is nothing to install for those. Firecrawl-shaped clients work by swapping the base URL.
- What are the rate limits and how is support handled?
- Rate limits scale with your tier, from the free tier up through Scale. Support is community and email on the lower tiers, with priority response on Pro and Scale. The status page tracks live availability.
Still have a question?
Run a scrape in the playground, or read the docs. The answer is usually one request away.