Package Information
Documentation
n8n-nodes-leapfun
This is an n8n community node for LeapFun AI, a text-to-speech service with four engine tiers, 100+ voices, 130+ languages and director's notes that steer emotion, pace and accent in plain English.
n8n is a fair-code licensed workflow automation platform.
Installation · Credentials · Operations · Usage · Example workflows · Compatibility · Resources
Installation
Follow the installation guide in the n8n community nodes documentation. Search for n8n-nodes-leapfun in Settings → Community Nodes.
Self-hosted n8n can install the node today. On n8n Cloud, community nodes appear once n8n has verified them; verification for this node is in review.
To use the node as a tool inside an AI Agent on a self-hosted instance, start n8n with N8N_COMMUNITY_PACKAGES_ALLOW_TOOL_USAGE=true. Note that an agent only receives the JSON part of the output (job ID, credits, engine); the audio file lands on the item's binary property, so route it to a storage node in the same workflow.
Credentials
- Sign in to app.leapfun.ai. Every plan includes the API, Free too; higher plans get higher rate limits, and commercial use of the audio starts on Starter.
- Open Settings → Developer API and click Create key. Keys start with
lf_sk_and are shown once. - In n8n, create a LeapFun API credential and paste the key. Leave Base URL at
https://api.leapfun.ai/v1.
The credential test calls GET /v1/usage, so a green check also confirms that your plan includes API access. API calls spend the same credits as the LeapFun app.
Operations
| Resource | Operation | What it does |
|---|---|---|
| Speech | Generate | Converts up to 5,000 characters into an MP3 or WAV file on the item's binary property (default data). The JSON output carries jobId, creditsUsed, engine, voice, format and ignoredParams. |
| Voice | Get Many | Lists preset voices and the cloned/designed voices in your workspace, with filters for engine, gender, language, search text and type. |
| Usage | Get | Returns credits used and remaining in the current billing period and your per-minute rate limit. |
Engines
| Engine | Best for | Billing |
|---|---|---|
| Ultra | Cinematic narration, ads, character work; director's notes | 1.5 credits per UTF-8 byte |
| Pro | Natural multilingual voices with emotion directives | 1 credit per character (or per byte on Ultra-shared voices) |
| Expressive | Expressive English voices; cloned voices speak 23 languages | 1 credit per character |
| Turbo | Fast, economical narration in 9 languages | 0.5 credits per character |
Pick the engine first; the Voice list is filtered to voices available on that engine. Ultra voices are also available on Pro at standard-tier billing.
Director's notes (Ultra and Pro)
Under Options you can set a Style (newscaster, vocal smile, whisper, storyteller, hype promo, deadpan), a Pace (natural, rapid fire, the drift, staccato), an Accent (American, British, Australian) and a free-text Director's Note of up to 300 characters, for example "warm and unhurried, like a late-night radio host". Options that a voice cannot honour are ignored and listed in ignoredParams rather than stopping the workflow.
Usage
A typical flow is Trigger → write or fetch text → LeapFun (Speech → Generate) → store the file. The audio lands on the binary property named in Put Output File in Field (default data), ready for Google Drive, Dropbox, S3, Telegram, or the Write Binary File node.
Error messages tell you what to do next:
| HTTP | Message |
|---|---|
| 401 | The API key was not accepted. Check the LeapFun credential. |
| 402 | Not enough credits for this request. Top up at app.leapfun.ai/billing. |
| 403 | LeapFun refused this request for your plan or workspace role. Check leapfun.ai/pricing or ask a workspace admin. |
| 413 | The text is longer than 5,000 characters. Split it into smaller items. |
| 429 | LeapFun rate limit reached. The node retries up to three times, honouring Retry-After, before surfacing this. |
| 502 / 503 | Engine or API busy. Retried up to three times; LeapFun refunds credits before answering 5xx, so retries never double-bill. |
Turn on Continue On Fail to receive these as JSON on the failing item instead of stopping the workflow.
Example workflows
Import any JSON from the examples folder via Workflow → Import from File:
rss-to-voiceover.json: RSS feed → OpenAI writes a 60-second script → LeapFun Ultra with a director's note → Google Drive.sheet-batch-narration.json: Google Sheets rows → LeapFun Pro per row → file back to Google Drive.telegram-text-to-voice.json: Telegram message text → LeapFun Turbo → voice reply.
Compatibility
Built and tested against n8n 1.100 and later with n8n-workflow 2.x. Requires Node.js 20.15 or newer for local development.
Resources
Development
npm install
npm run lint
npm run build
npm run dev # starts a local n8n with the node loaded (Docker required on n8n 3.x)
Releases are published from GitHub Actions with npm provenance (npm run release). Local npm publish is not used.