Exa, Firecrawl or Apify: Which Should My AI Agent Use?
I named all three on October 7, 2026 and didn't put them in order. The file on my session page splits them by the question you're asking before a call.
I didn't rank the three on the recording. The file on the session page gives each a job: recent news to Exa, a website and press page to Firecrawl, public social posts to Apify. Gumloop's documentation adds that web search and web fetch are built-in abilities where you pick the provider, and Exa and Firecrawl are on both lists (Checked on 2026-11-05). If one tool isn't connected, the file says to work from the other two.
You've got a call on your calendar and a prep brief to build. I named three research tools in the session, and you want to know which one to connect. The better question is what you want to know about the person before you dial.
Which tool goes with which prep question?
The instructions file on the session page, the one my chief of staff AI agent runs on, gives each tool a job:
| The question before the call | The tool the file names | Where I talked about it | |---|---|---| | What's been in the news about them? | Exa | 27:17 | | What's on their website and press page? | Firecrawl | 29:24 | | What are they posting in public? | Apify | 28:05 |
The sentence behind those rows:
From the file on the session page. Look up recent news on them with Exa, read their website and press page with Firecrawl, and check their public social posts with Apify.
What did I say about each one?
Exa first. I described it by comparing it to something you already use:
From the recording, 27:30. Exa is similar to Google, but it's built for agents.
Apify is the one for websites that don't have connectors, LinkedIn being my favorite example (28:21). I described it as a tool made for AI agents:
From the recording, 28:51. Apify is an agent native tool that fetches specific information from various platforms such as LinkedIn.
And I gave its limit in the same breath (29:18):
From the recording, 29:18. If it's a private page, it will not read it
Firecrawl is the one that goes through websites:
From the recording, 29:40. Firecrawl, similar to Exa and Apify, it is a system built to help agents to be able to process information on the web.
Did I name a winner?
Not across the three. I called all three tools I typically use (30:46). The one comparison I made was for a single job, prepping with LinkedIn:
From the recording, 29:01. If you're prepping and you're using LinkedIn to prepare for a meeting, Apify is one of the best tools for that.
Those rows are jobs. I didn't rank them.
What if one of the three isn't connected?
The same file covers it:
From the file on the session page. If one of the three isn't connected, work from the other two. If none are connected, tell me plainly you're working from general knowledge only, not a live lookup.
It also says what the three have in common:
From the file on the session page. Your research connectors read public information only. None of them log into a private account, and none of them post anything anywhere.
How do I choose?
Steps, using the file's order:
- 01
Write the question you'd ask before your next call: news, website and press page, or public posts.
- 02
Connect the tool the table gives for that question.
- 03
If you skip one, the file has the AI agent work from the other two.
What does the documentation add?
The builder in my session was Gumloop. On my relationship to Gumloop, I said this:
From the recording, 33:22. I don't work for Gumloop. I found this tool over the summer through Lenny's podcast, and I started using it and have since made myself an ambassador for them. I've created a relationship with Gumloop. I am not an employee.
More on that is in Do you work for Gumloop? (the answer from the session).
Checked on 2026-11-05 against Gumloop's AI agent documentation. In its Abilities table, Web Search lets you choose the provider from Exa, Parallel, Firecrawl or the model's native search, and Web Fetch lets you choose from Firecrawl, Parallel, Exa or Gumloop. The page doesn't mention Apify there. It also says the connectors you add define what the AI agent can see and do, and it tells you to pick the few your AI agent actually needs rather than connecting everything.
What goes wrong
A prep with no research tool connected. The AI agent has no live lookup, so the file tells it to say plainly that it's working from general knowledge only.
A private page. Apify reads public pages, and I told the group a private page won't be read (29:18).
I didn't connect Exa, Firecrawl or Apify on the recording.
Where this comes from
This article grows from the session I taught live with Women Applying AI on October 7, 2026. The quoted passages from the recording are my own words, taken verbatim. The passages marked as from the file on the session page come from the files published with that session. The steps, tables and checklists around them were drafted by my team's AI agents from those words and my published pages, and checked against the recording before publishing. Screens change, and platform facts that have a public source carry the dated line above.
Build Your First AI Agent in One Hour: the full session with the clips, the prompts and the transcript.
Related reading: Which Connectors Does an AI Agent Need?, What Is an AI Agent-Native Tool?, What Is an AI Agent Connector?.
Frequently asked questions
Which one do I connect first?
The one that answers your first question before the call. The file on the session page gives news to Exa, a website and press page to Firecrawl, and public social posts to Apify.
Did you say one of them is best?
Only for one job. For prepping with LinkedIn, I said Apify is one of the best tools for that (29:01).
Can it read a private LinkedIn page?
No. I said a private page won't be read (29:18).
What happens if I connect none of them?
The file tells the AI agent to say plainly that it's working from general knowledge only.
Build something that actually runs your workflow.
A focused, free 60-minute live session with Jacqueline. You build alongside her, on your own real task, and leave with an AI agent that is already running.