Skip to main content
The Lev8 API turns natural-language objectives into matching entities, contact details, researched answers, enriched records, or extracted public lists. Use the HTTP API from server-side products, or connect an AI client through the hosted MCP endpoint. You need an API Token to make requests. Create one from API Tokens and set it in your environment — see Authentication for a step-by-step guide.
The response includes a leads_search_id. Use it with GET /v1/entity-search/status to check progress and GET /v1/entity-search/fetch to read available results.

Quickstart

Create a task, check status, and fetch the first result batch.

Authentication

Create, store, and rotate an API key safely.

Entity Search

Review task fields, entity types, fetch parameters, and response objects.

Contact Search

Find an email address or phone number with a JSON request.

Deep Search

Answer an open-ended research question synchronously.

Enrich

Generate schema-driven fields for an existing entity.

Extract

Extract social, GitHub, or public web-page lists.

MCP server

Connect an AI client to typed Lev8 tools over Streamable HTTP.

Agent Skills

Install the Lev8 skill for Codex, Claude Code, or Cursor.

How it works

Authentication uses an Authorization: Bearer header. Entity Search reserves credits up front and settles against ready_count when the task completes. Idempotent replays return existing results without duplicate billing. The hosted MCP endpoint calls the same API and does not bypass scopes, concurrency, or billing.