Enrichment platform · Updated
Clay review (2026): is it worth the price?
Clay made table-based enrichment mainstream and still has the biggest provider marketplace. Two billing meters, a 50,000-row ceiling, and an API that runs what you built in the UI are the price of admission.
pipe0 sells data enrichment and competes with Clay in parts of this review. We say where Clay is better. How we review
Our score for Clay, the average of four criteria
The most complete enrichment table on the market, with 150+ providers and a huge template community, held back by a two-meter credit model and a learning curve measured in weeks.
Best for
- GTM engineers and RevOps teams who live in a table
- Agencies that reuse templates across clients
- Teams that want 150+ data providers under one bill
Skip it if
- You need tables larger than 50,000 rows
- You want agents or code to build the workflow, not just run it
- You need a bill you can forecast to the dollar
Templates help, but most teams need a specialist or an agency to build and debug tables that run in production.
Waterfalls across 150+ providers give it coverage no single database matches. Quality depends on the order you build.
Two meters that expire and top up on different rules, a 30% top-up premium, and BYOK lookups still billed in actions. Hard to plan and expensive at volume.
Agents can run functions a person built, 100 items at a time. They can't build or edit tables, and reading tables over the API is Enterprise-only.
Clay at a glance
- Founded
- 2017 in Brooklyn, New York. Headquartered in New York. Source
- Funding
- Series D of $115M at a $7.1B valuation, September 2026 Source
- Customers
- 17,000+, according to Clay (September 2026) Source
- Data providers
- 150+ in Clay's docs and pricing FAQ (its navigation says 200+) Source
- Rows per table
- 50,000 on paid plans, 200 on Free Source
- Seats
- Unlimited on every plan Source
- API and CLI
- Public since July 9, 2026, on all plans Source
- MCP clients
- Claude, ChatGPT, Copilot, Gemini, Grok, and Cursor Source
- Tables API
- Read-only, Enterprise plans only Source
What is Clay?
Clay is a spreadsheet where every column can call a data provider, an AI model, or another tool. You load a list of companies or people, add columns that find work emails, phone numbers, tech stack, or hiring signals, and push the results into a CRM or a sequencer. It has been doing this since 2017 and, more than anyone, made the enrichment table the default way GTM teams build lists.
The company raised $115 million at a $7.1 billion valuation in September 2026 and says it has more than 17,000 customers. That size shows in the product: Claygent for web research, signals, CRM enrichment, a workflow builder, and since July 2026 an API, a CLI, and an MCP server.
How Clay works
A Clay table is a program you can see. Each column is a step, each row runs through the steps, and each cell shows what came back. A typical build looks like this: import accounts from HubSpot, find three people per account by title, run a work-email waterfall that tries several providers in order, ask Claygent to summarize each company's last press release, and send the rows that pass a filter to a sequencer.
The waterfall is where Clay earns its reputation. You choose which providers to try and in what order, and Clay stops at the first one that answers. With 150+ providers in the marketplace, few lookups have nowhere left to go.
That flexibility comes with a cost in time. Waterfall order, run conditions, formulas, and column dependencies are all yours to get right. Teams that do well with Clay usually have one person who owns it, or they hire one of the many agencies that build Clay tables for a living.
What Clay costs in practice
Clay bills in two currencies. Data credits pay for data from Clay-managed providers. Actions pay for every step that runs: each enrichment, each AI prompt, each HTTP call. Since the March 2026 repricing, each plan's price is the sum of the two, and the pricing page shows both parts.
The meters behave differently, and that is the part to model before you buy. Actions expire at the end of each month. Monthly data credits roll over, but only up to twice the monthly amount. You can buy more data credits at a 30% premium, but you can't buy more actions; you move up a plan. A team that runs many cheap steps (formulas, AI prompts, HTTP calls) can burn through actions while credits pile up, and the reverse happens to teams that mostly buy data.
Two policies work in your favor. A lookup that finds nothing costs nothing, and bringing your own provider API key costs one action and no data credits. Clay's own calculator puts a work email at about 6 cents on the Launch plan billed annually, and a phone number at about 24 cents. Treat those as the floor: they assume the first provider answers.
Agents, the API, and the CLI
Clay released a public API and CLI on July 9, 2026, on every plan, with a plugin that bundles them with an MCP server for Claude, ChatGPT, Cursor, and others. It is a real step forward from webhooks.
The model behind it is that a person builds and an agent runs. The API runs functions and workflows that already exist in Clay, 1 to 100 items per call, and you poll until the results are ready. Reading tables over the API is limited to Enterprise and is read-only. Workflows, still in alpha, can now be edited from the CLI, which points to where Clay is heading. For now, an agent working with Clay is mostly a remote control for a table someone built by hand.
Clay vs pipe0
pipe0 has the same kind of table, with columns that run enrichments and waterfalls, but sheets hold up to 2 million rows instead of 50,000. Every sheet capability is also a REST API and an MCP server, so an agent can build the workflow, not only run it, and can do one-off lookups without creating a table at all. Waterfalls bill only the provider that answered, and we measure pipe0 as 3 to 10 times more cost efficient than Clay-style tools. Test that on your own list; the full comparison has the details.
Clay is ahead on the things that take years: the template library, the community, the agency ecosystem, and the number of providers in its marketplace. If your team already has a Clay expert and your lists fit in 50,000 rows, those are good reasons to stay.
Who should buy Clay
Buy Clay if one person on your team will own it, your lists are tens of thousands of rows rather than millions, and you value coverage across many providers over a predictable bill. Agencies get the most from it, because templates and the community carry across clients.
Look elsewhere if you want agents or code to build the workflow, if your lists outgrow 50,000 rows, or if the finance team needs a number they can forecast. For contact data only, a dedicated waterfall or a database like Apollo will be simpler.
Clay pros and cons
Pros
The largest provider marketplace
Clay lists 150+ data providers on one subscription, plus Claygent for web research. If a data source exists, it is probably a column in Clay already.
Templates and a real community
Clay has the largest template library and the most active community in the category. Agencies and freelancers sell Clay builds, so help is easy to hire.
Unlimited seats on every plan
Seats are free on every plan, including Free. You pay for usage, not for headcount, which suits teams where many people look and few people build.
You don't pay for misses
If an enrichment returns nothing, Clay charges neither data credits nor actions. BYOK lookups cost one action and no data credits.
Cons
Two meters that move separately
Data credits pay for data and actions pay for every step that runs. They are priced, topped up, and rolled over differently, so you can run out of one while the other sits unused.
A learning curve measured in weeks
Clay is closer to a low-code platform than a sales tool. Waterfall order, run conditions, and formulas all need someone who has done it before, and there is no sandbox for practice runs.
50,000 rows per table
Every plan stops at 50,000 rows per table (200 on Free). Larger lists get split across tables, or moved to Enterprise workarounds like passthrough tables.
The API runs what the UI built
The public API runs functions and workflows that already exist, 1 to 100 items per call. Reading tables over the API is Enterprise-only and read-only.
Clay pricing
Clay has a free plan. Paid plans start at $167/mo for Launch, billed annually ($185 billed monthly).
| Plan | Price | What you get |
|---|---|---|
| Free | $0 | 100 data credits and 500 actions a month. 200 rows per table, no phone enrichment, no CRM integrations. |
| Launch | $167/mo annual, $185/mo monthly | 2,500 data credits and 15,000 actions a month. Up to 50,000 rows per table, unlimited seats. |
| Growth | $446/mo annual, $495/mo monthly | 6,000 data credits and 40,000 actions a month, unlimited seats. |
| Enterprise | Custom, annual | 100,000+ data credits and 200,000+ actions. Tables API, passthrough tables, and bulk enrichment. |
- Clay's own calculator assumes 1 data credit and 3 actions per work email, and 5 data credits and 3 actions per phone number. On Launch, billed annually, that is about 6 cents per email and 24 cents per phone, before retries and validation failures.
- Actions don't roll over. Monthly data credits roll over up to twice the monthly amount; anything above that is dropped at renewal.
- Top-ups are data credits only, at a 30% premium over plan rates. Running out of actions means moving up a plan.
- A waterfall provider that returns an email which later fails validation is still charged.
Prices as of Oct 6, 2026, from Clay's pricing page.
Clay alternatives
| Alternative | Pick it when |
|---|---|
| pipe0 | You want Clay-style tables up to 2M rows, plus stateless API and MCP access where an agent can build the workflow, not just run it. |
| Apollo | You want one simple per-seat tool with its own database, sequences, and a dialer, and your market is mostly US tech. |
| FullEnrich | You only need emails and phone numbers from a waterfall, at a flat price per result, without building tables. |
| ZoomInfo | You are an enterprise buying a US data contract with intent data, and budget matters less than procurement. |
Side by side, feature by feature: pipe0 vs Clay.
Frequently asked questions
Is Clay worth it?
For a GTM engineer or agency that builds tables every day, yes: no other tool has as many data providers or templates. For a small sales team that wants contact data without learning a platform, the cost and the weeks of setup are hard to justify.
How much does Clay cost?
Clay has a free plan. Launch costs $167 a month billed annually or $185 billed monthly, Growth costs $446 or $495, and Enterprise is custom. Every plan has unlimited seats. Prices as of October 2026.
How much does Clay charge per email?
Clay doesn't publish a rate card. Its pricing calculator assumes 1 data credit and 3 actions per work email, which is about 6 cents on the Launch plan billed annually. A phone number is assumed at 5 data credits and 3 actions, about 24 cents. Real costs depend on the providers in your waterfall.
Does Clay have an API?
Yes. Since July 9, 2026, Clay has a public API, a CLI, and an MCP server on all plans. The API runs functions and workflows that were built in Clay, 1 to 100 items per call. Reading tables through the API is limited to Enterprise plans and is read-only.
How many rows can a Clay table hold?
50,000 rows per table on paid plans and 200 rows on the free plan. Enterprise customers work around the limit with passthrough tables and bulk enrichment.
Clay or Apollo: which should I choose?
Choose Apollo if you want one tool for finding contacts and sending sequences, and your market is mostly US tech. Choose Clay if you need coverage from many providers and custom enrichment logic, and someone on the team will own the build.
Does Clay charge for failed lookups?
No. An enrichment that returns no result costs no data credits and no actions. The exception is a waterfall provider that returns an email that later fails validation; that provider is still charged.
Next-gen enrichment & search.
Build revenue systems that scale. For humans, agents, and apps. Replace tools like Clay, n8n, Hightouch, etc.
Try pipe0