Clodo searches 1.3B profiles from a plain-English persona description, qualifies each match, and returns enriched profiles with verified email and phone for recruiting, expert sourcing and B2B prospecting.
Used this tool? Rate it
Used this tool? Rate it
Clodo is an AI people search engine: you describe the person you need in plain English, and it searches the open web plus proprietary datasets in real time, qualifies each match against that description, and hands back enriched profiles with verified contact details. It is operated by Astride Research, Inc., whose terms are governed by Delaware law; the company was founded in 2025 and came out of Y Combinator's Summer 2025 batch.
A contact database sells rows to filter; Clodo sells the act of finding someone matching a description no filter set can express, such as "compliance leaders at Series A-C fintechs nearing SEC requirements who are actively hiring compliance associates".
The homepage puts the searchable index at 1.3B profiles and names LinkedIn, GitHub, bioRxiv and PubMed among the sources feeding it. That mix says who it is built for: bioRxiv and PubMed are research-literature indexes, not B2B contact sources, which is why the vendor's examples lead with bench scientists rather than account executives.
The recruiting landing page states 1.2B+ profiles instead, and the homepage's own statistics band cites a third figure of 1.5B+, so vendor pages disagree by hundreds of millions of records with no dated methodology note on any of them. Treat index size as an order-of-magnitude claim, and ask for the current number during a demo if coverage depth drives your decision.
Natural language candidate sourcing works here because the input takes a persona description, not a query language, moving effort from boolean syntax to writing a specific enough sentence. Reported signals include active tool evaluation, hiring activity and budget authority — timing indicators filter-based tools usually lack or sell separately.
Contact enrichment is not a separate step you buy back later: every person Clodo returns, and every row of a list you upload, comes back with verified work email, mobile phone, LinkedIn URL, current title, company metadata and buying signals attached. A list held elsewhere runs down the same path rather than needing a second vendor.
Enrichment through the API is described as carrying sub-one-month freshness, which sets the expectation for how stale a returned email or title can be. That number matters, because "verified" alone says nothing about when verification happened.
The outbound layer sends under published rather than implied limits. Through the API a single request can draft a one-off email or a sequence of up to four steps for up to 100 contacts, and drafts require approval before they go out unless you deliberately opt into automatic sending. Sends run through a connected Google or Microsoft inbox rather than a shared vendor domain, and follow-ups stop on their own once a recipient replies.
Together those constraints describe a conservative design: the four-step ceiling rules out long drip sequences, approval-by-default stops unprompted sending in your name, and inbox-based sending puts deliverability on your own domain rather than a pooled one.
| Access surface | Status |
|---|---|
| REST API (search, enrichment, outreach) | Available on request |
| Claude / Cursor skill, all endpoints | Available today |
| Hosted MCP server | Roadmap |
| OpenAPI specification | Roadmap |
Agent access ships today as a drop-in skill for Claude and Cursor covering search, enrichment and outreach; a hosted MCP server and an OpenAPI spec are named as roadmap items, not shipped surfaces. A plan that assumes a machine-readable spec to generate clients from does not hold yet.
A sourcing run starts with the description, not the filters, so the quality of the output tracks how much qualifying detail you put into the sentence. The trade the product makes is explicit: no boolean strings, no filter dropdowns, no spreadsheets, and in exchange the ranked list comes back with verified emails, phones and the signals the engine picked up on the way. The vendor's worked examples read better as templates than as marketing.
The published examples cluster into distinct jobs, better read as a fit test than a feature list.
Each is a query filter interfaces handle badly. If your persona reduces to "VP of Sales, 50-200 employees, United States", this is not your problem.
The budget gate is the clearest fit signal the vendor publishes, and it filters harder than the feature list does. Check whether your team sits above the commercial floor before assessing search quality, because the intake assumes it does.
Two published facts do most of that filtering. There is no self-serve price list on clodo.ai; the single published figure sits inside the demo request form, where the stated floor is $1k per month or $12k per year. Before a conversation starts you are asked to pick a budget band of under $1k per month, $1k to $3k per month, or above $3k per month, which makes budget qualification part of the intake rather than the negotiation.
Good fit:
Poor fit:
Delivery is browser-first with an API layer behind it. Native integrations cover Salesforce, Dynamics, HubSpot, Bullhorn and Zoho, plus an API for anything custom, so sourced people can be pushed into an existing CRM or ATS instead of living in an export.
Stated infrastructure commitments are high rate limits, US-based hosting and high-availability SLAs, plus direct engineering contact instead of a ticket queue. US hosting is a platform property, not a configurable region, which matters if your data residency requirements point elsewhere.
Two surface-area changes matter. The Clodo Copilot Chrome extension was discontinued in March 2026, so the keystroke-level browser capture its old terms describe is no longer part of what you switch on today. Older writeups mentioning it describe a component that is gone.
Clodo has also absorbed people search traffic from Clado, a separate company retiring its own people search platform and API, which is worth knowing if you arrived from that migration. The names differ by one character and the businesses are distinct.
The vendor publishes exactly one number, inside the demo request form rather than on a pricing page. Every other price signal comes from third parties.
| Source | Stated entry price | Terms |
|---|---|---|
| Clodo demo form | $1,000 / month or $12,000 / year | Where engagements start |
| Third-party directory | $125 / month | Claimed Individual tier, 3,000 credits, 150 per run |
| Juicebox (independent roundup) | $99 / user / month | Billed annually |
| SeekOut (independent roundup) | $149 / month | Billed annually |
The sales motion expects a defined budget, not an exploratory trial, which is why the intake form asks for a band before your use case.
The two figures do not reconcile and should not be averaged. One AI-tool directory publishes a self-serve Individual tier at $125 per month with 3,000 credits and 150 credits per discovery run; that tier appears nowhere on the vendor's own site and sits an order of magnitude below the engagement floor the vendor states, so the two figures should be treated as unreconciled rather than as a range. A retired tier, an unpublished plan or an aggregator error would each explain it, and none of them can be settled from public pages alone.
One contract term is fixed whatever you agree and worth settling before signing: the refund policy. With annual billing it makes contract length a commitment rather than a discount, so treat evaluation as the only reversible window.
Published entry pricing in the wider sourcing category runs from $99 per user per month billed annually for Juicebox and $149 per month billed annually for SeekOut, both of which name a number in public where Clodo does not. That gap does not prove Clodo costs more per unit of value, since products meter differently, but any like-for-like comparison needs a sales call on one side and a published page on the other.
Its entry on a third-party product listing carries no user reviews at all, so no rated buyer sample exists to weigh against the vendor's own testimonials.
Several published terms change the risk profile of running outbound here, and matter more than any feature gap.
Accounts are restricted to adults, with the services stated as not intended for anyone under 18. Clodo says it does not sell personal information, and it separately secures consent to use data collected through the service, including user-generated content and interaction data, to train and improve its own machine learning models. Those statements answer different questions, and only the first is about onward sale.
Read the account-linking clause twice if you run multiple brands or a distributed team: it authorises suspension by association, not only by conduct. Two colleagues signing up separately from one office network fits the pattern, and the vendor decides alone whether that counts as abuse.
It takes a description rather than a filter set. A database returns rows matching criteria you already know how to express; Clodo qualifies each match against a written persona, including publication history, job movement or hiring activity most filter interfaces cannot represent.
One figure is published: engagements start at $1k per month or $12k per year, on the demo request form. There is no public price list, self-serve checkout or tier table. A third-party directory lists a $125 per month Individual plan absent from the vendor's own site, and the two cannot be reconciled from public information.
No. The terms state all payments are final and no refunds will be issued under any circumstances, listing partial use, dissatisfaction, account termination and service modifications as non-qualifying. Evaluate before committing.
No. Deliverability is explicitly not guaranteed and the terms place domain and IP reputation on the customer. Sequences send from your own Google or Microsoft inbox, so reputation damage accrues to your domain.
Three different figures appear across the vendor's own pages: 1.3B in the homepage source strip, 1.2B+ on the recruiting page, and 1.5B+ in the homepage statistics band. None carries a date or methodology note, so confirm the current number directly rather than quoting any of them.
Yes. The API covers search, enrichment and outreach, and agent access ships today as a drop-in skill for Claude and Cursor. A hosted MCP server and OpenAPI specification are near-term roadmap items, not shipped features.
Clodo states it does not sell personal information, and it separately secures consent to use data collected through the service, including user-generated content and interaction data, to train and improve its own machine learning models. Those are two different questions whose answers point in different directions.
On the compliance side the platform states CASA Tier 2 certification alongside SOC 2, GDPR, CCPA and DSGVO standards for how sourced data is handled. Those are vendor statements on a marketing page; ask for the underlying reports if your procurement process needs evidence rather than assertion.
