HeyGen Video 1
HeyGen's video model, ranked fourth in the text-to-video arena when we checked, which produces dialogue and sound in the same pass as the picture.
Updated November 9, 2026
A humpback whale gliding through golden clouds above a 1950s seaside town at dusk
At a glance
| Maker | HeyGen |
|---|---|
| Strengths in the app | Fast, Sound |
| Arena Elo (Artificial Analysis) | 1,152 |
| Typical wait | about 33 s |
| Who can use it | Paid plans |
| How it’s paid | Credits, from the first use |
| Privacy | Anonymized: NeverNo sends the request to its maker without your name, account or IP address. The maker still reads the text, applies its own rules and may keep it for a while under its own policy. |
Price
What the app charges, read from the same table the servers bill with. 1 credit = $0.01.
| 5-second clip | 45 credits ($0.45) |
|---|---|
| 10-second clip | 90 credits ($0.90) |
What it is
HeyGen Video 1 is a video model from HeyGen, offered in NeverNo's Video Studio on paid plans and paid in credits. It makes clips of 5 or 10 seconds in 16:9 or vertical 9:16, at 1344 by 768 pixels, from a text description or from a starting image. Unlike P-Video 2 Pro, Wan 3.0 or Seedance 2.5, it takes a starting image only: there is no ending image to set. The Studio tags it fast and audio, and it is the one model in the catalog whose dialogue and sound come in the same pass as the picture.
On the Artificial Analysis video arena, read on November 7, 2026, HeyGen Video 1 has an Elo of 1152 from text, fourth place in the arena at the time we added it, and 1123 from an image. For comparison, P-Video 2 Pro has 1128 and 1117, MiniMax H3 Max 1145 and 1115, and Veo 3.1 Fast 956 and 950. From text, only Wan 3.0 (1172) and Seedance 2.5 (1155) are higher among the Studio's models.
A typical 5 second clip takes about 33 seconds to render, as measured on Runware, our video provider. In our test it wrote a bakery sign correctly and had the baker wave at the camera, two things that are easy to get wrong. The Studio tags it for legible text, and when your description asks for written words, Auto gives a bonus to models carrying that tag.
It is anonymized. NeverNo sends the request to HeyGen without your name, account or IP address; HeyGen still reads the prompt and applies its own rules, and its content policy is not published, so some requests may be declined. The finished clip is stored in your browser, not on our servers. Think of it as a small casting tool. Write the person with a few visible traits, the room they stand in, the line they say and the gesture that goes with it, then check whether the lips, the voice and the background hum agree. If the voice sounds off for your scene, change the description of the speaker rather than the room. Short, natural sentences work best, and a single speaker keeps the timing tidy. For longer scripts, cut the text into several clips and assemble them in your editor. Trainers, shop assistants and language teachers can sketch a short welcome, a product tip or a pronunciation example, then compare takes to find the delivery that feels friendliest and clearest on camera. Rehearse the line aloud before typing it: if it trips your tongue, it will likely trip the performance too. Shorter sentences, plain vocabulary and a calm setting produce the steadiest results, and two or three versions of one scene will show which wording sounds most natural. Watch the mouth, the eyes and the hands first; if those three feel believable, the rest usually follows.
Strengths
Dialogue and sound generated in the same pass as the picture.
Arena Elo of 1152 from text and 1123 from an image (Artificial Analysis, read November 7, 2026).
Tagged for legible text; our sign test passed.
About 33 seconds for a typical 5 second clip.
Limits
Takes a starting image only; for a start and an end image, pick P-Video 2 Pro, Wan 3.0 or Seedance 2.5.
Anonymized, not private: HeyGen's content policy is not published, and some requests may be declined.
Needs a paid plan, and each clip is paid in credits.
When to choose it
Talking-head and character clips
A baker greeting customers, a guide welcoming visitors to a museum, a narrator introducing a product: describe the person, what they say and the setting, and the clip comes with the voice and room sound already in place.
Storefront and signage scenes
Because it renders words in the picture reliably, it suits scenes where a name matters: a cafe front, a poster in a station, a book cover held up to the camera.
Animating a portrait or an illustration
Start from a still character, add a sentence of speech and a gesture, and obtain a short animated greeting for an intro, a newsletter or a channel banner.
Prompt tips
Put the spoken line in quotation marks and name who says it and how, for example the baker says warmly, then waves.
Describe the room sound in a few words, such as oven fan humming, so the audio matches the scene.
One speaker and one gesture per clip gives the cleanest result; use 10 seconds only if the line is long.
Frequently asked questions
Does HeyGen Video 1 make dialogue?
Yes. Dialogue and sound are produced in the same pass as the picture, and the Studio tags the model audio.
Can I use an ending image?
No. It takes a starting image only. P-Video 2 Pro, Wan 3.0 and Seedance 2.5 accept both.
How fast is HeyGen Video 1?
About 33 seconds for a typical 5 second clip, measured on Runware, our video provider.
Is it private?
It is anonymized: NeverNo sends the request without your name, account or IP address, but HeyGen reads the prompt and applies its own rules. Clips are stored in your browser.
HeyGen Video 1 or Veo 3.1 Fast?
HeyGen Video 1 is rated higher on the arena (1152 versus 956 from text, read November 7, 2026) and accepts a starting image, which Veo 3.1 Fast does not. Veo 3.1 Fast, from Google, makes clips of 4, 6 or 8 seconds.