HeyGen Video 1.0 Image-to-Video: animate a still up to 2K
HeyGen Video 1.0 animates a first-frame image from a written scene. How to write the prompt in beats, which resolution to test at, and the frame it keeps.
Who is this model for? Someone who already has the picture. A product shot, a character you designed, a frame from a storyboard: HeyGen Video 1.0 Image-to-Video takes that image as the first frame and plays the scene forward from it, following a written description of what happens and how the camera moves. You do not get to redraw the opening; you get to decide what happens next. If you are still looking for the picture itself, start with a text-to-image model and come back.
What it gives you to work with
- The image. Required. It is the first frame, and the video keeps its aspect ratio, so a portrait photo makes a portrait clip.
- The prompt. Required. The field asks for the scene, the action and the camera movement, and it has room for a long, detailed description.
- Duration. From five to fifteen seconds, five by default.
- Resolution. 480p, 768p, 1080p or 2K, with 768p as the default.
- Prompt enhancement and seed, both under the advanced settings. Enhancement is off by default.
Write the scene in beats
Look at the example on the model page and you will notice it is not one long sentence. It is a tiny screenplay: where we are, what each person does, a pause, the moment that matters, then a separate paragraph for the camera and one for the mood. That shape works well here. Each beat is one thing the model has to show, in order, and the camera paragraph keeps movement instructions from getting tangled up with the action.
The ceramic teapot from the photo sits on a wooden table by a sunny window.
Steam begins to curl from the spout. A hand enters from the right and slowly pours tea into a small white cup.
The camera starts close on the spout, then pulls back to show the table, the cup and the window light.
Calm morning mood, warm natural light, soft shadows.The knight in the image stands at the edge of a cliff in strong wind.
His cloak lifts and snaps behind him. He turns his head toward the valley and rests a hand on the hilt of his sword.
The camera circles slowly from his side to a view over his shoulder, revealing the valley and a distant castle in the mist.
Epic fantasy, overcast light, dramatic scale.The storefront in the picture at dusk. The neon sign flickers on, a few people walk past, a cyclist crosses the frame from left to right.
Static camera on a tripod, no zoom.
City evening atmosphere, light rain on the pavement.Test cheap, finish sharp
The price grows with seconds and resolution, and you can see it on the Generate button before you confirm. So the sensible routine is to find the motion first and pay for the pixels last. Render your first tries at 480p and five seconds. Once the action reads well, note the seed of the take you like and render it again at 1080p or 2K, with the length you actually need. A new resolution may not reproduce the clip exactly, but a fixed seed and the same prompt tend to keep it close; the seed explainer shows why.
When to switch on prompt enhancement
Enhancement rewrites and expands your prompt before the render. It is useful when your description is one line and you want the model to fill in the details of the scene. If you have already written beats like the ones above, leave it off: you have made the decisions yourself, and a rewrite may add things you did not ask for. Of the two options, turbo is the quicker one and quality the more thorough.
The honest limitation
The first frame is a promise the model keeps at the start and then interprets. Faces, hands and small text drift the further the clip moves from that first picture, and a fifteen-second clip has a lot more room to drift than a five-second one. Large motions, a full turn of the body or a person walking out and back in, are where it shows first. For a person who has to stay recognisable, keep the motion small; the motion prompt glossary has the words for it. There is also no setting for sound or speech, so plan for the clip as picture only.
Open the studio, choose the model on its page and start from your best still.
Questions
Can HeyGen Video 1.0 change the aspect ratio of my image?
No. The video keeps the aspect ratio of the first-frame image, so crop the picture to the shape you need before you upload it.
How long can a HeyGen Video 1.0 image-to-video clip be?
From five to fifteen seconds. Five is the default, and short clips keep the subject closer to the original image.
Which resolution should I use?
Test at 480p to find the motion, then render the take you like again at 1080p or 2K. The price for each choice shows on the Generate button.
