Pick HeyGen Avatar IV when you have a photo and want that specific person on screen. Pick HeyGen Avatar V when you don't need your own likeness and want a transparent background, an SRT caption file or 4K. Both cost $0.15 per second, so the choice never touches your budget.
HeyGen Avatar IV is live on HiAPI. Open the model page, upload a portrait, type a script, and the person in the photo starts talking.
Avatar IV vs Avatar V at a glance
| Avatar IV | Avatar V | |
|---|---|---|
| Who appears on screen | Your own uploaded photo (image_url) | One of 497 preset avatars (avatar) |
| How it speaks | Script + 102 voices, or your audio | Same (the same 102 voices) |
| Resolution | 720p / 1080p | 720p / 1080p / 4k |
| Aspect ratios | 16:9, 9:16, 4:5, 5:4, 1:1, auto | Same |
| Background | Fills empty space when the frame differs from the photo (color or image) | Color or image background, remove_background strips the original, fit sets contain / cover |
| Transparent output | — | output_format: webm |
| Captions | — | caption: true returns a separate SRT file |
| Script length | 5,000 characters | 5,000 characters |
| Price | $0.15 / second | $0.15 / second |
The first row is the real decision: are you bringing the person, or picking one from a library? The rest is post-production extras that come with Avatar V.
Where does the presenter come from?
Avatar IV needs one photo. Whoever is in it becomes the presenter, with their clothes, hair and room carried over as-is. This 9-second explainer was rendered at 1080p, 16:9, with the Cute Chloe - Friendly voice:
Lip-sync, blinks and small nods all track the voice. The script:
Hi, I'm Lin. With one photo and a short script, I can present your product or record a lesson, with natural lip-sync and expressions.
The script is filled in; upload your photo and generate → Set the ratio to 16:9 and resolution to 1080p.
Avatar V skips the photo entirely. You pick a presenter from 497 presets. Here is Brandon Business Standing Front with the Jack Sterling - Broadcaster voice, 1080p, 16:9:
Photo guidelines only matter on the Avatar IV side. A front-facing solo shot with an unobstructed face, even lighting and a relaxed closed mouth works best. Illustrations and 3D characters work too, as long as the face is clear.
Same script, both models
Each pair uses the same script, the same voice, 1080p and 16:9, once on Avatar IV and once on Avatar V. Same voice and same words give the same length, so each pair also costs the same.
Pair 1: product intro (voice Cute Chloe - Friendly, 9 seconds, $1.35)
Hi, I'm Lin. With one photo and a short script, I can present your product or record a lesson, with natural lip-sync and expressions.
Avatar IV: the uploaded sweater portrait is the presenter. The room, hair and clothing match the photo; lips, blinks and small nods follow the voice.
Avatar V: the preset Abigail Office Front, a filmed presenter standing in front of a bookshelf and meeting table, gesturing as she speaks.
Generate with this script → The script is pre-filled; upload your own photo to generate.
Pair 2: client briefing (voice Dexter - Professional, about 8 seconds, $1.20)
Hi, I'm Daniel. Send me your quarterly numbers, and I'll walk your team through what changed and what to do next.
Avatar IV: a consultant's half-body photo with floor-to-ceiling windows and a skyline behind him; the suit and framing stay as photographed.
Avatar V: the preset Brandon Office Sitting Front, seated at a desk in an open-plan office with a laptop in front of him.
Generate with this script → The script is pre-filled; upload your own photo to generate.
Side by side: when the video has to show a specific person, use IV. When you just need a polished presenter to deliver the lines, pick one from Avatar V's presets.
How do they speak?
Identically. Write a script and choose from the same 102 preset voices, or drive the lip-sync with your own recording. Both accept up to 5,000 characters of script, in English or Chinese.
That means one script and one voice carry over unchanged when you switch models. Only the presenter field changes: Avatar IV takes a photo URL, Avatar V takes an avatar name.
How do backgrounds work?
On Avatar IV, the background is whatever the photo shows. background only comes into play when the output ratio differs from the photo and leaves empty space: pass #RRGGBB for a solid color, or an image URL to fill it. Here a landscape photo of a teacher sits in a 9:16 frame, with a chalkboard image filling the top and bottom, 11 seconds:
Hello everyone. Today we'll explain the Pythagorean theorem in five minutes. In a right triangle, the squares of the two legs add up to the square of the hypotenuse.
The script is filled in; upload your photo and generate → Set the ratio to 9:16 and choose an image background.
Avatar V goes further. remove_background strips the original backdrop so you can swap in a color or image, and fit decides whether the presenter is contained in full or covers the frame. output_format: webm gives you a transparent background, ready to layer over any shot in your editor.
Tip: Avatar V's live-action presets come with the backdrop they were filmed in, like the office behind Brandon above. If you need a uniform backdrop, render one short line first and check the background before you batch.
Captions and resolution
Set caption: true on Avatar V and you get an SRT file alongside the video. The picture stays clean, and you style the captions however you like in post. This vertical promo has captions on:
For resolution, Avatar IV offers 720p and 1080p; Avatar V adds 4k. Use Avatar V for big screens or 4K delivery. For mobile and web placements, 1080p is plenty.
How do the prices compare?
They're identical: $0.15 per second, regardless of resolution.
| Length | Avatar IV | Avatar V |
|---|---|---|
| 10 seconds | $1.50 | $1.50 |
| 30 seconds | $4.50 | $4.50 |
Both bill by the actual seconds of video generated, so cost depends only on how long the script runs. Full length-by-length pricing and monthly budgets are in HeyGen Avatar IV API pricing.
Which one fits your project?
| Use case | Pick |
|---|---|
| A brand spokesperson, instructor or founder appearing as themselves | Avatar IV |
| An illustration or 3D character you want to bring to life | Avatar IV |
| No suitable photo, and you just need a presenter | Avatar V |
| Layering the presenter into an edit with a transparent background | Avatar V |
| A separate caption file | Avatar V |
| 4K output | Avatar V |
FAQ
Do the two models cost the same? Yes. Both are $0.15 per second: $1.50 for 10 seconds, $4.50 for 30.
Can Avatar IV use my own voice recording?
Yes. Pass audio_url and the lip-sync and pauses follow your recording, up to 600 seconds. Avatar V supports audio-driven video too.
Does Avatar IV output transparent video or captions? For a transparent background or an SRT file, use Avatar V. Avatar IV's strength is putting your own photo on screen.
Which one handles cartoon or 3D characters? Avatar IV. Upload the character image; it just needs a clear face.
Bring your own character to life
Got an illustration or a 3D character? This chef script runs 8 seconds on Avatar IV:
Tonight's special is a slow-roasted tomato soup. It's simple, warm, and ready in twenty minutes. Let's cook!
The script is filled in; upload your character and generate → Prefer a preset presenter? Open the HeyGen Avatar V model page.
Related reading
- HeyGen Avatar IV model page: upload a photo and generate online
- HeyGen Avatar IV API pricing: per-second rates and monthly budgets
- HeyGen Avatar IV examples: 7 scripts with real videos
- HeyGen Avatar IV API tutorial: from photo to talking video
- HeyGen Avatar IV API docs
- HeyGen Avatar V API docs: avatar, remove_background, caption and more
