HeyGen TTS

Badge: Direct
Node type: audio:heygen_tts
Category: audio

Description

Speak a script in a voice you cloned with HeyGen Voice Clone, on HeyGen Voice 1. Wire the clone into voice_id, or pick one of your HeyGen voices.

Pricing

• Cost: 1.5 credits per estimated second of speech, 5 minimum: a 3-second line is 5 credits, a 15-second read 23, a 50-second read 75. Estimated from the text (characters / 15), never re-priced from the audio.

Canvas ports

These appear as port handles on the left side of the node.

ID	Label	Details
text	Text	TEXT (required) — Plain prose, up to 750 characters per run, priced by length. No SSML or break tags.
voice_id	Voice	TEXT (required) — Connect the HeyGen Voice Clone node, or pick one of your HeyGen voices.

Sidebar config

These render as form fields in the right-side config panel when the node is selected.

ID	Label	Details
model	Model	TEXT · options: heygen-voice-1, heygen-voice-1-turbo
language	Language	TEXT · options: auto, ar, be, bg, ca, +35 — Language of the text. Auto lets HeyGen detect it from the text.
expressiveness_boost	Expressiveness	NUMBER · 0…1 — 0 to 1. 0.5 keeps closest to the cloned voice (HeyGen Studio default); 1 is more animated.

Outputs

ID	Label	Type
audio	Audio	AUDIO

How to use it

Wire a HeyGen Voice Clone node's Voice ID into Voice, or pick one of your saved HeyGen voices in the sidebar. Wire a script into Text and click Run Model. The audio player shows up on the node when the run finishes. The file is a WAV stored on Wireflow's CDN.

This node speaks only voices you cloned with HeyGen Voice Clone. Another account's voice, or a HeyGen voice that was not cloned in Wireflow, is refused before anything is charged.

What a run costs

Priced by length: 1.5 credits per estimated second of speech, with a 5-credit minimum. A 3-second line costs 5 credits, a 15-second read 23, a 50-second read 75. The estimate is the character count divided by 15, where digits and symbols count 5 each and Chinese, Japanese and Korean characters count 3, because they take longer to say. The canvas shows the price of the text you wired from a Text Input. When the words are not known before the run (they come from a model, or an API call, automation, template or blueprint wire replaces the Text Input), it shows the most a run can cost (75 credits), and the charge follows the text that actually runs.

A run reads up to 750 characters, about 50 seconds of speech. Longer text is refused before anything is charged, so split a long script into parts and join the audio.

HeyGen bills speech by the minute of audio it makes, not by character. The price follows HeyGen's published HeyGen Voice 1 speech rate.

Settings

• Model: HeyGen Voice 1, or Voice 1 Turbo for lower latency.
• Expressiveness defaults to 0.5, which keeps closest to the cloned voice. 1 is more animated.
• Language is detected from the text. Set it when the text is short or mixes languages.

Plain prose only: no SSML, no break tags.

---

Auto-generated from the Wireflow node registry.