Company Intelligence Data Built for AI.
PredictLeads sells structured company-intelligence data rather than a tool: news events, job openings, technology detections, financing events, SEC filings, products, connections, GitHub repositories and website evolution, all keyed to a company graph the vendor puts at 120M+ companies. Delivery is deliberately plural - REST API, webhooks, flat files, and a hosted MCP server - and the pitch is aimed squarely at agents and autonomous systems as consumers, not just at humans.
| Plan | Price | Included | Notable |
|---|---|---|---|
| 100 API credits (Free) | $0 | Up to 100 free API calls per month | Full API access at trial volume, No commitment |
| 101–5,000 API calls | $0.04 per API credit, $40 minimum monthly charge | — | Pay-as-you-go, No hidden fees, no commitments |
| 5,001–10,000 API calls | $0.02 per API credit | — | Priority support |
| 10,001–100,000 API calls | $0.01 per API credit | — | 24/7 support |
| 100,001–500,000 API calls | $0.004 per API credit | — | 24/7 support |
| 500,001+ API calls | $0.002 per API credit | — | 24/7 support, Contact the vendor above this volume |
| Enterprise Access | Contact sales | Full flat-file access to all datasets, full historic data from 2015, unlimited companies tracked | Data exports via AWS S3, GCS or SFTP, Real-time webhook access to all datasets, Delta exports |
Checked 2026-08-07 against their pricing page. Vendors change pricing without notice.
PredictLeads is a signal layer. The documentation is organised dataset by dataset - each with its own data model, endpoints and webhooks - which tells you the intended pattern: pick the signals your motion cares about, follow the companies you care about, and receive changes rather than poll for them. The distinctive design choice is the deleted-record datasets (Job Openings Deleted, News Events Deleted), which exist so downstream systems can retract as well as add. The MCP framing is not marketing dressing either: the integrations page lists Claude Desktop and ChatGPT as first-class consumers alongside the automation platforms.
Teams building something on top of company signals - GTM engineers wiring trigger-based outbound in Clay, n8n or Cargo, data teams enriching a warehouse, investment researchers tracking funding and hiring, and agent builders who want a company-data tool their model can call directly.
This is raw data with a credit meter, not a workflow: below 5,000 API calls a month you pay a $40 minimum monthly charge regardless of usage, credits do not roll over and expire at the end of each billing period, and following a company costs a credit per month whether or not it produces a signal - so occasional, bursty or exploratory use is the worst-value pattern here. It is also not a contact database: the datasets are company-level (firmographics, signals, technographics), so a team that needs people, emails and phone numbers needs a different vendor alongside this one. Flat-file and full-history access is Enterprise-only and gated behind a sales conversation.
No curated rival set covers this tool yet, so these are simply other tools in the same architectural layer and product form. Treat it as a starting point, not a shortlist.
Pay-as-you-go credits with volume tiers, from $0.04 per credit at 101 - 5,000 calls down to $0.002 per credit above 500,000. There is a $40 minimum monthly charge once you leave the free tier, and 100 free API calls a month below that. Enterprise flat-file access is quoted separately.
Per the vendor's pricing FAQ: every API request costs 1 credit regardless of endpoint or how many records come back, with one exception - Discovery endpoints are billed per result returned. Each company you follow costs 1 credit per month.
No. The pricing page states credits do not roll over and unused credits expire at the end of each billing period.
Yes - the vendor's llms.txt lists a hosted MCP server as a delivery method alongside REST, webhooks and flat files, and the integrations page shows it wired to Claude Desktop and ChatGPT and listed in the Smithery, PulseMCP, MCP Servers and MCP Market directories. We did not retrieve the server itself or an enumerated tool list.
Company-level events: the News Events dataset alone is categorised into 37 event types including product launches, partnerships, funding rounds, expansions, leadership changes and hiring plans, alongside separate datasets for job openings, technology detections, financing events, SEC filings, products and inter-company connections.
Not from these datasets. Everything documented is company-level: firmographics, signals, technographics, postings and filings. There is no person dataset in the documentation we retrieved. Pair it with a contact-data vendor if you need people.
Follow companies and receive signals on a webhook URL - most datasets document their own API Webhooks section, and there are dedicated deleted-record datasets (Job Openings Deleted, News Events Deleted) so your system can retract records as well as add them.
Treat the spec as claimed but unverified from our side.
The Enterprise flat-file tier advertises full historic data from 2015, point-in-time, with delta exports. We did not retrieve any statement about history depth available through the API tiers.
This page tells you whether PredictLeads fits. Cheetah builds and runs the system it goes into — enrichment, sequencing, CRM and the joins between them.
Book a call →The directory stays free. There is nothing gated behind this.