Firecrawl Alexandria vs Treg: Which People Search Layer Should Your AI Agent Use?

Posted on Nov 3 • Originally published at apidog.com

Firecrawl Alexandria vs Treg comes down to how you want to pay and who picks the provider. Alexandria puts 430 providers, including 12 People sources such as Apollo, People Data Labs and OpenSanctions, behind your Firecrawl key, billed in plan credits. Treg puts 123 providers behind one token, routes each people call to a provider for you, and bills a prepaid balance at the provider's own rate, with misses at $0.

Alexandria is a provider library you call by name. Treg is a router that picks a provider per call.

Alexandria is wider; Treg is deeper on contact data.

The Alexandria library lists 430 providers and 2,461 capabilities across 20 categories. Its People category holds 12 providers: four commercial enrichment sources (Apollo, FullEnrich, People Data Labs, Data Legion), Particle for podcast speakers, and seven public-record sources, including OpenSanctions, the UK Sanctions List, the CMS NPI Registry and US Senate financial disclosures.

Those public records are the clearest gap. If your agent screens names against sanctions lists or checks an NPI number, Alexandria has a named tool; we didn't find one on Treg's people-search pages.

Treg's people-search page quotes 1B+ contacts across 123 providers, with Apollo, Hunter, Tomba and People Data Labs among them. Its routed capabilities cover the prospecting loop: treg.people.search, treg.people.email.find, treg.people.email.verify and treg.companies.enrich.

Our Firecrawl Alexandria explainer covers all 12 People providers in detail.

The biggest design difference is who chooses the provider.

Alexandria: you pick. Per the Alexandria docs, your agent discovers tools with POST /v2/search and sources: ["alexandria"], inspects one with find-tools, then runs it through POST /v2/scrape with an alexandria object:

curl -s https://api.firecrawl.dev/v2/scrape \ -H "Authorization: Bearer $FIRECRAWL_API_KEY " \ -H "Content-Type: application/json" \ -d '{"alexandria": {"provider": "apollo", "capability": "people/search", "options": {"person_titles": ["VP of Sales"], "q_organization_domains_list": ["example.com"], "per_page": 10}}}' Enter fullscreen mode Exit fullscreen mode Results come back in alexandria[] with the payload under data. Some providers first return THIRD_PARTY_DATA_TERMS_REQUIRED with a requiresAction.url, and an org admin has to accept that provider's terms.

Treg: the router picks. You call a routed capability and Treg chooses the provider, cheapest first, and moves to the next one on a miss or a 429/5xx:

curl -X POST "https://treg.to/call/treg.people.email.find" \ -H "X-Treg-[REDACTED CREDENTIAL] \ -H " content-type: application/json " \ -H " X-Treg-Route-Max-Cost: 0.05 " \ -d '{" full_name ":" Jane Doe "," domain ":" example.com "}' -D - Enter fullscreen mode Exit fullscreen mode The charge comes back in X-Treg-Cost-Micro (micro-USD). Direct tools such as apollo.people.search exist too, and don't fail over.

Want results from one known source? Alexandria's provider field pins it. Care more about the answer than its source? Treg's routing writes the fallback logic for you.

Alexandria spends the same credits as the rest of Firecrawl. Discovery is free. Execution costs the capability's listed credits. The Government and Research indexes are free up to a daily limit set by your plan. A firecrawl_feedback call within 20 minutes refunds 1 credit, capped at 10 per website and 100 per team per UTC day.

Plans from the Firecrawl pricing page, monthly billing:

Cost per credit is price divided by credits. Paid plans add credits in $5 increments: 1,000 on Hobby, 2,000 on Standard, 2,500 on Growth and 5,000 on Scale.

On November 3, the provider pages showed Apollo people/search at 0 credits and People Data Labs person/enrich at 60 credits. One PDL enrichment is $0.228 on Hobby and $0.0594 on Standard.

Treg has no subscription. It meters every call from a prepaid balance at the provider's rate with 0% markup. Misses on per-success providers bill $0.00. Your team's own provider keys are used first and never billed. New teams get $1.00 of credit once, and the cheapest email-find on the catalog starts at $0.0005.

Alexandria's credit price drops as your plan grows, which suits a team already spending Firecrawl credits. Treg's bill tracks answers, not plan size.

The task: find the VP of Sales at 10 company domains and get a verified work email for each. We didn't run it; these are listed prices from November 3.

Run both on 10 of your own domains and compare cost per verified email.

Firecrawl's test measures answer quality; Treg's measure email accuracy and cost. Neither compares Alexandria to Treg. Our Clay alternatives breakdown goes through the 100-lead table in more depth.

Your agent's workflow depends on a response shape staying put and a price staying under a cap. A regression suite in Apidog CLI catches a renamed field or a pricier route before your prospect list does. You write one prompt; the agent runs the commands.

Set up the CLI and its agent skill once:

npm install -g apidog-cli@latest apidog auth login --with-token <YOUR_APIDOG_TOKEN> npx -y skills add https://apidog.com # pick Claude Code, Codex or Cursor when prompted Enter fullscreen mode Exit fullscreen mode Then give your agent this [REDACTED PROMPT] Write an OpenAPI 3 file with two Firecrawl calls (POST /v2/search with sources ["alexandria"], POST /v2/scrape with provider "apollo", capability "people/search", per_page 5) and one Treg call (POST /call/treg.people.email.find). Firecrawl sends [REDACTED CREDENTIAL]earer {{FIRECRAWL_API_KEY}}; Treg sends X-Treg-[REDACTED CREDENTIAL] and X-Treg-Route-Max-Cost: 0.05. Import it into Apidog project, create "Firecrawl" and "Treg" environments, and write test cases that assert status 200, data.alexandria.data present with no error, and X-Treg-Cost-Micro at or below 50000.

Run each case against its environment and summarize any failures.

The agent validates each payload and runs commands like these: Enter fullscreen mode Exit fullscreen mode bash apidog import --project --format openapi --file./people-search-layers.json apidog environment create Firecrawl --project --base-url https://api.firecrawl.dev apidog environment create Treg --project --base-url https://treg.to apidog cli-schema get test-case-create apidog cli-schema validate test-case-create --file./alexandria-apollo-case.json apidog test-case create --project --file./alexandria-apollo-case.json apidog run --test-case --project -e \ --env-var "FIRECRAWL_API_KEY=$FIRECRAWL_API_KEY" -r cli,junit apidog run --test-case --project -e \ --env-var "TREG_TOKEN=$TREG_TOKEN" -r cli,junit