HeyGen Video · Clipia

HeyGen Video videos where the character speaks

Describe the scene and put the line in quotes — HeyGen's model films a clip where the character says your words and the setting sounds around them. Start from text, bring a photo to life or build the scene from image references.

Create with HeyGen Video Watch examples

clip length: 5–15 s
resolution: 768p · 1080p
in every clip: sound and speech

In short

What HeyGen Video is

HeyGen Video is a video model by HeyGen. It films clips of 5 to 15 seconds with sound built in: the character's lines, the voice and the ambient sound come from the text prompt, so no separate voice-over is needed. Describe the scene, upload a photo as the first frame or add references: up to 9 images, up to 3 clips and up to 3 audio recordings. Resolution is 768p or 1080p, with six aspect ratios from 21:9 to 9:16.

text, photo, references: three ways to set the scene
up to 9 images: references in one scene
6 aspect ratios: from 21:9 to 9:16
sound and speech: no separate voice-over

Text to video · speech in Russian

A chef in a restaurant kitchen

A clip from a single description: the chef tastes the sauce, nods and from the fourth second speaks to the camera in Russian — "The secret is simple: less salt, more patience". The lips match the words, pans sizzle and dishes clink behind him.

How to set up the scene

Describe the place, the character and the action, put the line in quotes and name the language and the delivery: "says in Russian in a warm confident voice". List the ambient sounds separately — sizzling pans, steam, clinking dishes.

Try this mode

Photo to video · speech in English

A photo that says hello

The first frame of the clip is the source portrait. The man comes to life, smiles and says in English: "A minute ago this was just a photo. Now I can say hello." Blinking, slight head movement and the quiet room tone come from the prompt too.

How to set up the scene

The photo becomes the first frame, so the clip takes its aspect ratio from the image. In the prompt describe what the person does and says after that frame, and add the background sound: "soft office room tone, sunlight from the window".

Try this mode

References · character and product

Perfume in the character's hands

Two references — a portrait of a woman and a photo of a bottle. The model built a living-room scene: she turns the bottle to the light and says in Spanish "Huele a verano junto al mar". The bottle keeps its square shape, gold collar and amber colour.

How to set up the scene

Refer to the references in order: "the woman from the first reference image holds the bottle from the second". List what must stay exact — shape, material, colour — then add light and camera movement.

Try this mode

Open the model in the video studio. Limits and cost are shown in the editor.

Create with HeyGen Video

Parameter	Value	Note
Duration	5–15 seconds	Whole seconds
Resolution	768p or 1080p	Chosen before you start
Modes	Text, photo, references	The photo becomes the first frame; references set the characters, product and place
Sound	Speech and ambient sound	Generated by the model from the prompt; in reference mode you can attach up to 3 audio recordings as a sample
Aspect ratio	21:9, 16:9, 4:3, 1:1, 3:4, 9:16	In photo mode the ratio comes from the image; reference mode also has an automatic ratio
Inputs	Text, one photo or references: up to 9 images, 3 clips and 3 audio recordings	A clip can be the only reference; an audio recording only goes with an image or a clip

Text to video

The scene, the character, the line and the sound from one description. Six aspect ratios — from wide 21:9 to vertical 9:16.

Try this mode

Photo to video

The image becomes the first frame: the person in the photo comes to life and says what the prompt says.

Try this mode

Scene from references

Up to 9 images of characters, a product and a place, up to 3 clips and up to 3 audio recordings. The model brings them together in one scene.

Try this mode

Sound and speech in every clip

The model creates the lines, the voice and the ambient sound itself. No separate voice-over or sound editing needed.

Try this mode

Parameter	HeyGen Video	OmniHuman 2.0
What you provide	Text, a photo or references: images, clips and audio recordings	A photo and a speech recording
Voice in the clip	Generated by the model from the prompt	Your uploaded recording
Length and quality	5–15 seconds, 768p or 1080p	4–60 seconds, 720p

Choose a mode

Text only, a photo as the first frame, or image references of characters, a product and a place.

Describe the scene and the line

The place, the action, the line in quotes with the language and delivery, the ambient sounds.

Set the parameters

Duration of 5–15 seconds, 768p or 1080p resolution and the aspect ratio.

Prepare a scene description with a line, a photo or references.

Create with HeyGen Video

A 5-second clip in 768p costs 5 credits, a 15-second one 25 credits. A 5-second clip in 1080p costs 35 credits. Sound and speech are included.

The price is the same for all three modes: text, photo and references.

Put the line in quotes

Write the phrase in quotes and name the language in words: "says in Spanish". In the examples the model spoke Russian, English and Spanish lines word for word.

Sound in a separate sentence

List the ambient sounds after the line: "pans sizzle, steam rises, dishes clink". This keeps the sound of the scene apart from the description of the action.

Match the duration to the line

In the examples a phrase of up to 13 words fitted into 8 seconds. Split long text into several clips.

Short social videos

A scene with a spoken line from text alone, in vertical 9:16 for Reels, Shorts and clips.

A photo that greets

A person or character in the picture says a couple of phrases: an invitation, a greeting, an introduction.

A product in the character's hands

A portrait and a product photo as references turn into an ad with a line in the language you need.

Before sending a request, check that the model is available in the catalog and get a cost estimate for the selected parameters.

Clipia documentation

FAQ

Is HeyGen Video available in Clipia?

Yes. The model is open in the video studio in three modes — text, photo and references. The video studio starts with it by default, and the buttons on this page lead straight to it.

How do the three modes differ?

From text the model invents the whole scene. In photo mode the image becomes the first frame and the clip takes its aspect ratio from it. In reference mode you can upload up to 9 images of characters, a product and a place, up to 3 clips and up to 3 audio recordings — the model brings them together in one scene.

How long is a clip and in what quality?

From 5 to 15 seconds, in whole seconds. Resolution is 768p or 1080p; the aspect ratio is 21:9, 16:9, 4:3, 1:1, 3:4 or 9:16.

How much does a generation cost?

A 5-second clip in 768p costs 5 credits, a 15-second one 25 credits; a 5-second clip in 1080p costs 35 credits. Sound and speech are included. The exact amount is shown in the editor before you start.

Are the demos on this page real generations?

Yes. The three clips were made with HeyGen Video, 8 seconds each in 1080p, with no editing or added voice-over: the speech and sound come from the model. The portraits were generated by AI and the voices are synthetic — there are no real people in the clips. The clips are compressed for the page.

Is HeyGen Video in the Clipia API?

Yes, when the model is listed in the Clipia catalog with the active status. Check the parameters and the cost before creating a task.

Create a video

Film a clip with speech in HeyGen Video

Describe the scene and the line or upload a photo. The cost is shown before you start.

Create with HeyGen Video

HeyGen Video demo

Watch the clip and explore its motion, detail, and camera work.

Watch demo

Open HeyGen Video in the editor

HeyGen Video — русская версия

HeyGen Video · Clipia

HeyGen Video videos where the character speaks

Create with HeyGen Video Watch examples

clip length: 5–15 s
resolution: 768p · 1080p
in every clip: sound and speech

In short