# PredictLeads — how to use (mcp.ai)

Connect your PredictLeads account and use 9 tools for AI sales tools straight from your AI agent. Connect with your own API key. PredictLeads provides company intelligence on hiring, technologies, financing, news, products, business connections, and other growth signals.

## Option A — via MCP (recommended)
Remote MCP endpoint (HTTP, streamable): `https://api.mcp.ai/p_predict_leads?ms=1787291160000`
Add it as a custom/remote MCP connector in your client (Claude, Cursor, VS Code…), then authenticate when prompted. Once connected, ask the agent to use the server's tools (e.g. `predict_leads_discover_companies`).

## Option B — via direct REST API
Base URL: `https://api.mcp.ai/api/predict_leads`
Auth: `Authorization: Bearer sk_live_…` — create a workspace API key at https://mcp.ai/settings/api-keys
Discover endpoints: `GET https://api.mcp.ai/api/predict_leads/_endpoints`

### Endpoints
- `POST https://api.mcp.ai/api/predict_leads/discover/companies` — Find companies using firmographic filters. Discovery costs one credit per returned record, so the default limit is 5. Deterministic upstream 4xx responses also consume credits and are not retried.
  - body: { page?: integer, limit?: integer, sizes?: string[], industry?: string, location?: string, naics_codes?: string[], revenue_range_low?: integer, revenue_range_high?: integer }
- `POST https://api.mcp.ai/api/predict_leads/discover/intelligence` — Discover one page of jobs, technology adoption, news, financing, investor portfolio connections, products, or startup-platform posts across companies. Select exactly one dataset. Each returned record 
  - body: { page?: integer, limit?: integer, title?: string, dataset: string, location?: string, date_from?: string, seniority?: string, date_until?: string, onet_codes?: string[], post_types?: string[], active_only?: boolean, company_sizes?: string[], news_categories?: string[], last_seen_at_from?: string, last_seen_at_until?: string, product_categories?: string[], product_subcategories?: string[], financing_types_normalized?: string[], technology_id_or_fuzzy_name?: string }
- `POST https://api.mcp.ai/api/predict_leads/get/api/subscription` — Check PredictLeads subscription status, monthly credit quota, and credits used. This endpoint is free and should be called before or during credit-sensitive workflows.
- `POST https://api.mcp.ai/api/predict_leads/get/company` — Retrieve one company by PredictLeads UUID or bare domain. Each upstream response costs 1 credit, including deterministic 4xx failures, so malformed identifiers are rejected locally and 4xx errors shou
  - body: { id_or_domain: string }
- `POST https://api.mcp.ai/api/predict_leads/get/company/intelligence` — Retrieve one page of a company's jobs, technology detections, news, financing, SEC filings, connections, website changes, GitHub repositories, or products. Select exactly one dataset. Each call costs 
  - body: { page?: integer, limit?: integer, dataset: string, location?: string, date_from?: string, date_until?: string, not_closed?: boolean, active_only?: boolean, job_categories?: string[], news_categories?: string[], last_seen_at_from?: string, last_seen_at_until?: string, product_categories?: string[], with_location_only?: boolean, company_id_or_domain: string, connection_categories?: string[], product_subcategories?: string[], with_description_only?: boolean }
- `POST https://api.mcp.ai/api/predict_leads/get/intelligence/record` — Retrieve one job opening, extended technology detection, technology, news event, SEC filing, or product by identifier. Select exactly one record type. Each call costs 1 credit; malformed identifiers a
  - body: { identifier: string, record_type: string }
- `POST https://api.mcp.ai/api/predict_leads/get/similar/companies` — Find one page of companies similar to a specified PredictLeads company UUID or domain. Each returned record costs 1 credit, so the default limit is 5; deterministic 4xx responses are not retried.
  - body: { page?: integer, limit?: integer, company_id_or_domain: string }
- `POST https://api.mcp.ai/api/predict_leads/list/followed/companies` — List one page of companies already followed for PredictLeads webhook updates. This is read-only and does not change follow state. The operation costs 1 credit per response; deterministic 4xx failures 
  - body: { page?: integer, limit?: integer }
- `POST https://api.mcp.ai/api/predict_leads/search/technologies` — Search PredictLeads' tracked technology catalog by fuzzy name. This unusually expensive endpoint costs 3 API credits per returned record, so the default limit of 3 can cost up to 9 credits. Check API 
  - body: { page?: integer, limit?: integer, order_by?: string, fuzzy_name?: string }

## Example prompts
- "What can I do in PredictLeads?"
- "Show me a summary of my PredictLeads account"

## More
- Page: https://mcp.ai/predict_leads
- Agent spec (llms.txt): https://mcp.ai/predict_leads/llms.txt
- Postman collection: https://mcp.ai/predict_leads/postman.json