Two tools — search and fetch — usable with no key, plus a separate authenticated endpoint for deep research.
LABSCO SUMMARY
Passing what you are trying to find out, alongside the search terms, lets the server select excerpts against the goal rather than against keyword overlap. That is the difference between a snippet that happens to contain the word and one that answers the question. Splitting fast search from authenticated deep research is the other sensible division: the cheap path stays open and anonymous.
Parallel publishes two endpoints. The search one lists two tools without credentials; a separate Task endpoint handles deep research and enrichment and always needs credentials.
Nothing for search; an API key as a Bearer token raises the rate limits. To sign in with OAuth instead, use the separate address https://search.parallel.ai/mcp-oauth, which refuses anonymous requests. The Task endpoint needs an API key or OAuth. One-click install links exist for Cursor, VS Code, LM Studio and Goose.
Paste a URL — add the endpoint to your client
USE CASES
State the objective alongside the queries so the excerpts are selected against it.
Fetch it when the excerpts are not enough — the documented order.
Use the authenticated Task endpoint and poll for the markdown result.
Counted once with only startup setup. Other versions, modes or optional credentials may expose a different list.
web_fetchPurpose: Fetch and extract relevant content from specific web URLs. Use only when web_search excerpts are insufficient for the task at hand.
Ideal Use Cases:
web_searchPurpose: Perform web searches and return LLM-friendly results, including excerpts that are usually sufficient to answer directly without a follow-up fetch.
Ideal Use Cases:
search_queries can be issued in a single callBEFORE YOU INSTALL
2 tools when we connected.
CONTEXT COST
Rated by the size of the 2 tool definitions that ride along on every turn, not by how many there are
INSTALLATION
Add the endpoint to your client.
https://search.parallel.ai/mcpMCP servers
OctagonAI
parallel-web
HughesCuit