n8n-nodes-glasser

Premium data for n8n workflows and AI Agents through one Glasser Key: find people, company intelligence, SEO research, web research, social research and market data, backed by Apollo, People Data Labs, Hunter, BuiltWith, DataForSEO, Semrush, Ahrefs, Serper, Exa, ScrapeCreators, RentCast and more. The node says what data it wants; Glasser decides which provider serves it. Pay per call, no signup at each vendor.

One node, one credential. usableAsTool is on, so the same node works as a step in a workflow and as a tool under an AI Agent.

Resources and operations

Each operation shows only the inputs it requires. Where the API accepts one of several identifiers, a selector picks it: Lookup By (LinkedIn URL, Email, or Name and Company) for Enrich a person, Search By (Address, City and State, or ZIP Code) for Property Records and listings. Optional inputs are under Filters (Search people), Additional Fields (Country) and Options.

Options → Provider exists on every resource. Auto (the default) lets Glasser pick the source for the operation and fall back to the next one on a provider error; naming a provider forces it.

How a call works

1. The node sends the parameters its form shows, flattened, to POST https://api.glasser.ai/v1/solutions/gtm/<resource> with a fresh idempotency key. Comma-separated fields become arrays; the Lookup By / Search By selectors are not sent; nothing else is changed.
2. Glasser resolves (operation, provider) to one endpoint, translates the fields into that endpoint's native input, runs it, and under Auto moves to the next source on a provider error. It never falls back on an empty answer: "no result" is an answer.
3. The output item is the run exactly as the Glasser API returns it: provider and endpoint say who served the call, input is what was sent, output is the provider's own payload, charge_usd is the exact charge as a decimal string, run_url opens it in the console. An async run is polled until it is terminal (up to 180 seconds).

The node holds no routing table and no state. Its parameters are generated from the Glasser OpenAPI document (scripts/gen-properties.mjs): field names, types, enums and descriptions come from the contract, and the generator only adds the form layout — which inputs each operation requires and which are optional.

Credentials

1. Sign in at https://app.glasser.ai, open Keys, create a Key (it starts with gl_). The workspace needs a balance: runs are prepaid and charged per call.
2. In n8n, add a Glasser API credential and paste the Key. The credential test calls GET /v1/balance, which is free.

Example workflow

examples/glasser-demo.json is a workflow to import (Workflow menu → Import from file): it searches people at a domain, finds a work email, searches the web and reads a domain rating. Pick your Glasser API credential on each node and execute.

Installation

Follow the installation guide in the n8n community nodes documentation. Package name: n8n-nodes-glasser.

Usage notes

• Each call is one paid run at the routed endpoint's published price. Under Auto a provider error can lead to a second run at the next source; each run is billed under its own terms and all of them are visible in the Glasser console.
• Two indicators, not one. A run's status and the provider's response are separate. A COMPLETED run whose provider answered 404 ("person not found") is a normal outcome, charged per the endpoint's clauses (charge_basis.clause).
• Errors are the API's own envelope. A refused request (unknown country, a filter the named provider cannot apply, a provider that does not serve the operation) is raised with Glasser's error code and message; nothing was run and nothing was charged.
• Money is an exact decimal string ("0.0005"), never a float.

Development

npm install
npm run gen                # regenerate nodes/Glasser/properties.ts from https://glasser.ai/docs/openapi.json
npm run lint
npm run build
npm run dev                # local n8n with the node loaded

Support

• Source repository: https://github.com/glasser-ai/n8n-nodes-glasser
• Glasser documentation: https://glasser.ai/docs
• Console: https://app.glasser.ai
• Contact: support@glasser.ai

License

MIT