HeyGen Video is the row for a 5–15 second music-video shot that already has sound. 480P or 720P, from text, a first frame, or up to nine reference images. In SunoMV a 5-second draft is 40 credits at 480P and 60 at 720P. Our 1 October 2026 probe returned a 5-second 480P clip with sound, 832×480.
HeyGen Video is a finished-looking short take: picture and sound together, optional reference images, 40 credits for 5 seconds at 480P. It will not pin a first frame and a last frame, and it will not do native 2K. If you are still throwing away ideas, MiniMax H3 Max is the iteration row. If you already know the face and want the chorus take to include sound, pick this one.
A lyric shot that arrives silent still needs a bed, a room, and a voice. This row generates the picture and the sound together, and the sound cannot be switched off.
Duration is 5 to 15 seconds. There is no 3-second or 4-second option. That is one lyric line or one chorus phrase, not a full song in a single generation.
The picker charges 8 credits per second at 480P and 12 per second at 720P. A 5-second shot is 40 credits at 480P and 60 at 720P. Draft the song at 480P, then restage the chorus at 720P.
Those are the two resolutions in the picker. 720P is the 768p picture. A 1080P request snaps down to 720P. Native 2K and 4K are a different row.
Text, a first frame, or a stack of references. There is no first-and-last-frame control on this model.
Start from a sentence, from one still, or from reference mode: up to 9 images, and you can also attach video and audio. Reference mode is how you keep a face across lyric lines.
You cannot pin the opening picture and the closing picture and ask this model to walk between them. That control lives on MiniMax H3 Max and Gemini Omni 1.1, not here.
Hailuo 03 / MiniMax H3 is the heavier video model: native 2K, lip-sync, multi-reference. HeyGen Video is the cheaper shot that already includes sound. They are different picker rows.
The model is already in the video picker. You do not apply for a separate key, and you do not leave the music-video flow to use it.
Open the create flow, choose HeyGen Video, set 5–15 seconds and 480P or 720P. The homepage model list is the same catalog, so the name you see there is the name in the picker.
Describe the scene, start from a still, or pass reference images when the singer has to stay the same person from verse to chorus.
Drop each clip on the lyric timeline, preview vertical, square or widescreen, and export when the chorus cuts clean. Previewing the assembly is free.
Nothing to install. The model is already in the SunoMV video picker and on the homepage model list.
Open create, choose HeyGen Video, set 5–15 seconds and 480P (draft, 40 credits) or 720P (keepers, 60 credits). Ignore 1080P — it snaps down to 720P.
Write the scene, start from a still, or add up to 9 reference images when the singer has to stay the same person. Do not look for a first-and-last-frame toggle — this model does not have one.
Lay the clips on the lyric timeline. Preview free, then export vertical, square or widescreen.
No video model is universally better — only better fitted. These are the jobs this row is for.
The clip comes back with sound. You are not generating a silent picture and then hunting for a room tone.
Reference mode takes up to 9 images. Use it when verse and chorus have to be the same singer, not a new person every bar.
40 credits per 5 seconds at 480P is the iteration price on this row. Confirm the shot, then restage the keeper at 720P.
Same music-video flow. Different picker rows. Click through instead of guessing.
5–15s, 480P/720P, sound included, up to 9 reference images. 40 credits per 5s at 480P.
Faster drafts with first-and-last-frame. No reference-to-video. 100 credits per 5s at 480P.
Start from a Suno link, then switch the shot model to HeyGen Video.
HeyGen Video is a text-to-video, image-to-video and reference-to-video model. Clips last 5–15 seconds at 480P or 720P and include sound. It is selected in SunoMV under that name for short music-video shots, not for native 2K or first-and-last-frame cuts.
A HeyGen Video clip is 5 to 15 seconds, in whole seconds. Shorter than 5 seconds is not a setting. A full song is many of these clips stacked on a timeline, not one generation.
Reference-to-video means the clip is generated from uploaded images, and optionally video or audio, instead of from text alone. HeyGen Video accepts up to 9 images in that mode, which is how a music video keeps one face across shots.
Capability limits match the SunoMV picker. Credits follow the published list price. The clip specs are our own probe on 1 October 2026.
HeyGen announced HeyGen Video on 30 September 2026: finished-looking video, with an October rate of $0.01 per second described as half of list, and a base the company calls a post-trained MiniMax H3.
HeyGen — announcement, 30 September 2026 ↗The public catalog lists text-to-video, image-to-video and reference-to-video, durations of 5–15 seconds, and resolutions of 480p and 768p.
HeyGen Video catalog ↗Published per-second list price: $0.02 at 480p and $0.03 at 768p for a text or image generation; reference generation is double. SunoMV credits are 8/s and 12/s from that list, not from the October promotion.
OpenRouter — heygen/heygen-video-1 ↗Our probe on 1 October 2026: text-to-video, 5 seconds, 480P. The finished clip is 832×480, about 5.3 seconds, with sound. The invoice was $0.0495, the October promotional rate.
SunoMV — measured clip, 1 October 2026 ↗FAQ'S
Ask us anything!
5–15 seconds, 480P or 720P, sound included. Draft at 40 credits, keep the chorus at 60. Preview the full-song assembly free, then export.
/icon.png){kind=icon}
