Integration · Company data

Connect Claude & ChatGPT to Parallel Web Systems

Give your AI the Parallel web APIs. It can search the web and get ranked excerpts, pull the content out of specific pages, and start longer research tasks that return structured answers. It can also find lists of companies or people that match your criteria and set up monitors that watch the web on a schedule.

Connect Parallel Web Systems free

Free plan, no credit card. Takes about a minute.

Try asking

“Search the web for recent news about battery recycling in the EU.”
The AI runs a web search with your queries and returns ranked results with excerpts.
“Pull out the pricing section from these three product pages.”
It extracts the pages and focuses the excerpts on the objective you gave.
“Research who founded Acme Robotics and when, and give me sources.”
It starts a research task run, checks its status and fetches the output when it is done.

What your AI can do

Search and extract

Run web searches with ranked, excerpted results and extract the content of pages you name, focused on what you need.

Research tasks

Start single task runs or task groups for many inputs, follow their status and read the finished output.

FindAll and monitors

Discover companies or people that match your criteria, and create, update, list and stop web monitors.

“Find 20 logistics software companies based in the Nordics.”
It starts a FindAll run with your conditions and returns the matched entities.
“Research these 50 companies and fill in headcount and HQ city for each.”
It creates a task group, adds one run per company and lists the runs with their outputs.
“Watch for news about our competitor every day.”
It creates a monitor with the settings and frequency you choose, and lists the events it finds.

20 tools for Parallel Web Systems

These tools are switched on when you connect. You can switch any of them off, or require your approval before it runs.

add_task_group_runsChanges dataAdd research runs to a task group. Pass `taskgroup_id` and `inputs` (a list of run inputs), optionally with a `default_task_spec`.
cancel_findall_runDeletesStop a running FindAll run by `findall_id`. This cannot be undone.
cancel_monitorDeletesStop a web monitor permanently by `monitor_id`. This cannot be undone.
create_findall_runChanges dataStart a FindAll run that discovers entities (such as companies or people) on the web matching your criteria. Requires `objective`, `entity_type`, `generator`, `match_limit` and `match_conditions`.
create_monitorChanges dataCreate a monitor that watches the web on a schedule. Requires `type`, `settings` (what to watch) and `frequency`.
create_task_groupChanges dataCreate a task group to hold many research runs that are processed together. Returns the `taskgroup_id` for add_task_group_runs.
create_task_runChanges dataStart a web research task. Pass `input` (a question as text or a JSON object) and `processor` (the Parallel processor tier to use); optionally a `task_spec` describing the output. Runs are billed per processor tier; poll with get_task_run.
extract_web_pagesChanges dataFetch and extract the content of one or more web pages. Pass `urls`, and optionally an `objective` to focus the excerpts on what you need.
get_findall_resultsGet the matched entities found by a FindAll run by `findall_id`.
get_findall_runGet the status of a FindAll run by `findall_id`.
get_monitorGet one web monitor's settings and status by `monitor_id`.
get_task_groupGet a task group's status and run counts by `taskgroup_id`.
get_task_runGet the status of a research task run; pass the run's ID (from create_task_run) as `task_id`.
get_task_run_resultGet the finished output of a research task run by `run_id`; optionally wait up to `timeout` seconds for it to complete.
list_monitor_eventsList the events (changes found) for a monitor by `monitor_id`.
list_monitorsList your web monitors, optionally filtered by `type` or `status`.
list_task_group_runsList the runs in a task group by `taskgroup_id`, optionally filtered by `status` and including inputs and outputs.
search_entitiesChanges dataQuickly find entities of an `entity_type` (such as companies or people) that match an `objective`, up to `match_limit` results.
update_monitorChanges dataChange a monitor's settings, frequency or processor by `monitor_id`. Only the fields you pass change.
web_searchChanges dataSearch the web and get ranked, excerpted results. Pass `search_queries` (a list of queries) and optionally an `objective` in plain words and `max_chars_total`.

Set up in three steps

1
Pick the app
Create a free PipMCP account and choose Parallel Web Systems from the app list.
2
Paste your key
Paste a Parallel API key. Sign in at platform.parallel.ai, create an API key and copy it.
3
Add the link to your AI
You get a personal MCP link. Add it to Claude, ChatGPT or Cursor:
1. Click your name, then Settings › Connectors › Add custom connector.
2. Paste your link as the Remote MCP server URL.
3. Switch it on from the + menu in a chat.

Questions

What can the AI do in Parallel Web Systems?

It can search the web, extract web pages, start and read research task runs and task groups, run FindAll searches and quick entity searches, and list, create, update and cancel monitors and read their events. Cancelling a FindAll run or a monitor is permanent.

Does the AI see my Parallel Web Systems credentials?

No. Your Parallel API key is encrypted at rest and never shown to the AI. After you save it, it is not shown again, not even to you. The AI only sees the results of the tools it calls.

Can I control what the AI is allowed to do?

Yes. You choose which tools are switched on, so you can for example allow search and extract only. Starting paid research runs and cancelling runs or monitors can require your approval before they run. Every tool call is logged.

Does it work with ChatGPT?

Yes. In ChatGPT go to Settings › Apps & Connectors › Advanced and turn on Developer mode, then add your PipMCP link. Developer mode needs a paid ChatGPT plan: Plus, Pro, Business or Enterprise. The same link also works in Claude (Settings › Connectors › Add custom connector), Cursor and other MCP clients.

What does it cost?

PipMCP has a free plan with no credit card required. Paid plans bill per completed task. You connect your own Parallel account, so anything Parallel charges for searches and task runs is between you and Parallel.

Can I choose the processor for a research run?

Yes. A research task run takes your question plus the Parallel processor tier to use, so you can tell the AI which one you want. Runs use your own Parallel account. If you want to check each run before it starts, set create_task_run and create_findall_run to require your approval.