Give an agent the live web.
Search mid-conversation and get results that are already scraped, not a list of links to go and fetch.
What it costs
- The search itself
- 1 credit
- Each result scraped
- 1 credit
- A search that returns nothing
- Free
[ The problem ]
A list of links is not an answer
Most search APIs hand back titles, URLs and a snippet. The agent then has to fetch each one itself, which is where it meets the blocks, the JavaScript and the HTML you did not want it reasoning over.
- Snippets are too short to answer from and too long to ignore
- Fetching each result yourself means handling blocks, rendering and retries in agent code
- Raw HTML burns context on markup the model then has to see past
Step 01
One call searches and scrapes. Results come back ranked, each already reduced to Markdown you can cite.
curl -X POST https://api.hydrafetch.com/v1/web/search \
-H "X-API-Key: $HYDRAFETCH_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"query": "retrieval augmented generation evaluation",
"limit": 5,
"scrapeResults": true
}'Step 02
Or skip HTTP entirely. Connect over MCP and the agent gets scrape, map, search, extract and brand as tools it can call on its own.
claude mcp add --scope user --transport http hydrafetch https://api.hydrafetch.com/mcp \
--header "Authorization: Bearer $HYDRAFETCH_API_KEY"[ Worth knowing ]
Connecting over MCP costs exactly the same as calling the matching endpoint over HTTP.
Set scrapeResults to false when the agent only needs to know what exists, and pay for the search alone.
Other use cases
RAG pipelines
Crawl a whole site on a schedule and pipe clean Markdown straight into your embeddings.
Structured extraction
A JSON schema in, typed records out, and the passage each value was drawn from.
Brand data
Logos that work on light and dark, the real colours ranked by how the site uses them, fonts and socials.
[ Start ]
Clean web data is one call away.
250 free credits, no card required. Failures are never billed.
