n8n-nodes-stealthgpt

This is an n8n community node. It lets you use StealthGPT in your n8n workflows.

StealthGPT turns prompts and existing drafts into natural, human, publish-ready long-form content via API. Use the full content pipeline (research → draft → optional fact-check → humanize → optional citations/images → final markdown) or just the humanization step on text you already have. The output is designed to read as natural and human, and to hold up against AI-content detectors and editorial quality filters.

n8n is a fair-code licensed workflow automation platform.

• Installation
• Operations
• Credentials
• How it works (async runs)
• Example workflows
• Billing & credits
• Resources
• Support
• License
• Changelog

Installation

Follow the installation guide in the n8n community nodes documentation.

In n8n: go to Settings → Community Nodes → Install, enter n8n-nodes-stealthgpt, and confirm.

Operations

The node exposes two resources, each with a Start Run and a Get Run Status operation.

Content Generation

Runs the full StealthGPT content pipeline.

Start Run fields:

• Preset — Academic, SEO, or Social.
• Prompt — the topic or brief to generate from (required).
• Platform — LinkedIn or Medium (shown only for the Social preset).
• Enable Fact Check — run an additional fact-checking step (default off).
• Enable Image Generation — generate images for the content (default off).
• Wait for Completion — poll until the run finishes and return the result (default on).
• Poll Interval (Seconds) / Max Wait (Seconds) — control polling when waiting (defaults 5s / 600s).
• Options — Idempotency Key, Webhook URL, Webhook Secret.

Get Run Status fields:

• Run ID — the ID returned when the run was started (content generation run IDs are prefixed with run_).

Text Humanization

Transforms existing text into natural, human-quality writing.

Start Run fields:

• Text — the text to humanize (required).
• Quality Mode — Quality or Fast (default Quality).
• Model — Heavy or Lite (default Heavy).
• Output Format — Text or Markdown (default Text).
• Wait for Completion — poll until the run finishes and return the result (default on).
• Poll Interval (Seconds) / Max Wait (Seconds) — control polling when waiting (defaults 5s / 600s).
• Options — Idempotency Key, Webhook URL, Webhook Secret.

Get Run Status fields:

• Run ID — the ID returned when the run was started.

The completed humanization payload includes howLikelyToBeDetected, a human-likeness score from 0–100 where higher means more human.

Credentials

You need a StealthGPT API key.

1. Create an account at stealthgpt.ai.
2. Get your API key from https://www.stealthgpt.ai/stealthapi.
3. In n8n, create a StealthGPT API credential and paste the key into API Token.
4. Leave Base URL as https://www.stealthgpt.ai unless instructed otherwise.

The credential is authenticated with the api-token header and tested against the balance endpoint. Your key stays in your n8n instance.

How it works (async runs)

StealthGPT runs are asynchronous. Starting a run returns a runId immediately; a full pipeline run can take several minutes.

• Wait for Completion = on (default): the node starts the run and polls the status endpoint every Poll Interval seconds, up to Max Wait seconds, then returns the terminal payload. If the run ends as failed or cancelled, the node throws an error with the API's message. If it doesn't finish in time, the node throws a timeout error that includes the runId so you can fetch the result later.
• Wait for Completion = off: the node returns { runId, status, statusUrl } immediately. Wire your own polling (a Wait node + Get Run Status) or use a webhook to continue when the run finishes.

Use the Options → Idempotency Key to make start calls safe to retry without creating duplicate runs or charges. You can optionally provide Webhook URL / Webhook Secret to have StealthGPT call back on completion; see the API docs for the signature scheme.

Example workflows

RSS → SEO article → WordPress draft

1. RSS Read trigger.
2. StealthGPT → Content Generation → Start Run, Preset SEO, Prompt built from the feed item, Wait for Completion on.
3. WordPress → create a draft post from result (markdown).

Webhook → Humanize → Google Docs

1. Webhook trigger receiving the source text.
2. StealthGPT → Text Humanization → Start Run, Output Format Markdown, Wait for Completion on.
3. Google Docs → insert result.

Long-running run with manual polling

1. StealthGPT → Start Run with Wait for Completion off → returns runId.
2. Wait node (e.g. 30s).
3. StealthGPT → Get Run Status with the runId.
4. IF status is completed → continue; otherwise loop back to the Wait node.

Billing & credits

Runs consume StealthGPT credits. Completed runs return billing metadata such as creditsSpent, remainingCredits, and billingMode. Using an Idempotency Key prevents duplicate charges on safe retries. Failed and cancelled runs behave predictably per the API contract. Manage credits and billing at stealthgpt.ai.

Resources

• StealthGPT API documentation
• Get an API key
• n8n community nodes documentation

Support

• Help Center: stealthgpt.zendesk.com
• Issues: github.com/StealthGPT/n8n-nodes-stealthgpt/issues

License

MIT

Changelog

See CHANGELOG.md.

Compatibility

Requires n8n with Node.js 20.15 or newer. Built and tested against the n8n nodes API version 1.