How does credit-based B2B data pricing work for AI agents?

When an AI agent can search a B2B database on your behalf, the obvious question is cost. What stops it from running a broad search and spending your whole balance? Enricho answers that with a simple model: free previews, approval before spending, and 1 credit per record you actually receive.

Why does pricing matter more when an agent does the searching?

In a traditional prospect database, you click the export button yourself. With an agent, the search happens on your behalf, often in several steps. That makes two things important:

1. Predictable unit cost. You should know what each result costs before you ask for it.
2. A checkpoint before money moves. The agent should show you what it found and wait for a yes.

Enricho is built around both.

What is free?

These actions cost nothing:

Previews share a daily allowance per account, which is larger on paid plans. The details are in the credits and limits docs.

How does the preview-then-approve flow work?

Every list-building request follows the same steps:

1. You describe the target. For example, directors of operations at logistics companies in Melbourne.
2. The agent previews. It runs a free preview and gets a sample of up to 10 records without contact details.
3. The agent shows its work. It tells you which filters it used and shows the sample.
4. You refine or approve. Adjust the targeting as many times as you like, or say how many full records to fetch.
5. The agent fetches one page. Each returned record costs 1 credit. It offers the next page instead of fetching more automatically.

Here is what that looks like in a chat:

Use Enricho to find heads of finance at companies in Sydney with
201-500 employees. Preview first.

The sample looks right. Fetch 30 full records.

The second message is your approval. Thirty returned records cost 30 credits.

Why don’t previews show the total number of matches?

Enricho does not expose exact match totals. A preview is a quality check: it tells you whether the titles, companies and locations look right, not how large the audience is. A sample of 10 does not mean only 10 people match, and it does not promise thousands either. If you want more, ask for the next page.

What happens if you don’t have enough credits?

If your balance cannot cover the whole page you asked for, no records are returned and nothing is deducted. The agent can ask for a smaller page, or you can add credits.

You can also ask at any time:

How many Enricho credits do I have left, and when do they refill?

The agent calls account_usage and reports your plan, credits left and refill date.

How much do the plans cost?

Credit packs cost $20 for 10,000 credits. Pack credits never expire and are used after your monthly credits. Monthly credits refill on the 1st and do not roll over. You can cancel any time, and the Free plan needs no card. Full details are on the pricing page.

Can you upgrade from inside the chat?

Yes. Ask the agent:

Upgrade my Enricho account to Pro.

The agent calls account_upgrade and returns a secure checkout link. Nothing is charged until you complete the payment. To add credits without changing plans, ask for a credit pack and the agent calls account_buy_credits. To update your card, download invoices or cancel, ask for the billing portal. Payments are processed by Dodo Payments as Merchant of Record.

How do you keep costs low?

1. Use narrow filters. Combine seniority, job title, industry, company size and city rather than searching broadly.
2. Require the contact channel you need. Adding has email or has phone avoids paying for records you cannot use.
3. Exclude what you already have. Pass existing customers and contacts as exclusions in the full search.
4. Preview until it looks right. Refining is free; only full records cost credits.
5. Ask for smaller pages. Start with 20 records, review them, then fetch more.
6. Ask the agent to be budget-aware. For example: "Check my Enricho credits first, then search with the narrowest filters you can."

How does this compare with other pricing models?

Traditional prospect databases often price by seat or by annual contract, with export limits on top. Per-record credits suit agent workflows because cost follows results: a failed search or an unmatched enrichment costs nothing, and you approve each paid step.

Where to go next

• New to the concept? Read what an MCP server for B2B data is.
• Ready to build a list? See how to build a B2B lead list in Claude or how to find B2B leads in ChatGPT.
• Enriching an existing list? See how to enrich LinkedIn profiles, emails or domains.
• Questions about volume or billing? Contact the team.