Hedra alternative
Hedra alternative: make any character talk, on the model you pick
Hedra made talking characters from a still and a voice famous. Pixel Dojo runs three avatar models for the same job, plus text to speech for the voice, and shows the price of each before you generate.
We made one 3D style character and one voice line, then handed both to three avatar models. This page explains how audio driven characters work, what each model did with them, and how the plans compare.
600 credits ($37.50) per clip. Included in every plan, from $10/month.
Already a member? Sign in
By Brian Lovett · Updated
How audio driven talking characters work
A talking character model takes three things: a still image with a clear face, an audio track, and an optional text prompt. The audio drives the mouth, the jaw and the timing of the head moves. The prompt steers the rest, such as gestures, mood and how much the body moves.
Hedra describes its Character 3 model this way: image, text and audio in, speech driven video out, for talking avatars, marketing clips and storytelling.
The still matters most. A front facing, well lit portrait with the mouth closed gives the model a clean face to work from, and stylized art works as well as photos. Clean audio matters next, because every syllable becomes a mouth shape.
More talking characters, and the stills behind them
Every clip and image below was made on Pixel Dojo with the prompt shown.
A painted character on P-Video Avatar, voiced from a typed script in the same run.
The source portrait on Seedream 5. Any illustration with a clear face can be the start.
Painterly fantasy illustration portrait of an elderly wizard tutor in his seventies with a long white beard, a deep blue robe with brass buttons and half moon spectacles, sitting at a cluttered desk of books and candles in a round tower study, looking at the viewer with a kind smile, warm candlelight, chest up framing.
A different route: Veo 3.1 Lite writes the speech, the face and the room in one clip, with no avatar step.
Medium close up of a podcast host in her thirties with short auburn hair and a mustard cardigan leaning into a broadcast microphone in a small studio with acoustic panels and a warm desk lamp, she smiles and says: "Today we are talking about why everyone suddenly wants to learn pottery." Sound: her voice close and clear, faint room tone.
The character still on Nano Banana 2.1 that all three avatar models animated.
3D animated film style character portrait of a cheerful mountain hut cook in her fifties with rosy cheeks, a grey braid, round glasses and a flour dusted red checked apron, standing in a warm wooden alpine kitchen with copper pots behind her, facing the camera, chest up, mouth closed with a gentle smile, soft warm key light, stylized but detailed.
One hut cook, one voice line, three avatar models
We made the cook on Nano Banana 2.1, voiced her welcome line with text to speech for 0.5 credits, then gave the same image, audio and prompt to three models. Each tile lists its settings and credits.
The cheerful hut cook greets her guests, speaking warmly to the camera with lively eyebrows, small nods and a friendly smile
The cheerful hut cook greets her guests, speaking warmly to the camera with lively eyebrows, small nods and a friendly smile
- OmniHuman gave the most acting, a full wave and the most expressive face, and it was the most expensive of the three at 50 credits.
- Kling Avatar was close behind for 20 credits, with a friendly hand gesture and natural nods.
- P-Video Avatar was the calmest and the cheapest at 10 credits. It suits explainers where the face should talk without distracting gestures.
- All three kept the 3D style, the glasses and the braid intact from the first frame to the last.
Hedra plans vs Pixel Dojo
Everything in the Hedra column comes from its pricing, Character 3 and home pages, read on October 15, 2026.
| At a glance | Hedra | Pixel Dojo |
|---|---|---|
| Monthly plans | Basic $20 for 2,000 credits, Pro $50 for 7,200, Ultra $100 for 18,000, Teams $75 for 14,400 | Lite $10 a month for 160 credits, Pro $25 for 500 credits, Pro Max $50 for 1,100 credits |
| Talking character price | Character 3: 2.5 cents a second at 540p, 5 cents at 720p, 6.25 cents at 1080p | Each avatar model page shows its credit price before you generate |
| Talking character models | Hedra Character 3, with a still image and audio required | OmniHuman, Kling Avatar and P-Video Avatar, with text to speech for the voice |
| Speed | Basic is listed with slower generations, Pro faster, Ultra fastest | The same models on every plan |
| Developers | Hedra API v3 and an MCP server | A public API with docs generated from each model's spec, a hosted MCP server and API credit packs from $5 |
| Commercial use | Commercial use on every paid plan | Commercial use on every plan |
Deep dive: choosing an avatar model for the job
For a character who should act, with gestures and big expressions, start with OmniHuman. For a friendly spokesperson clip at a lower price, Kling Avatar is the middle ground. For training videos, explainers and long scripts where the face just needs to talk, P-Video Avatar is the steady, low cost choice, and it can voice a typed script itself.
If you do not have a character yet, make one on Nano Banana 2.1 or Seedream 5, then save it in Character Studio so the same face can appear in later scenes by @name. When the scene needs the room to move too, Veo 3.1 can generate the speech and the shot together, as in the podcast clip above.
When Hedra is the better pick
Hedra says Character 3 runs up to 10 minutes in one generation, at up to 1080p, in seven aspect ratios. If you need one long, unbroken talking take, that is a real strength.
If you want to compare several avatar models on the same face and voice, or move on to films and ads with the same character, we think Pixel Dojo is the more flexible home.
What a talking character costs on Pixel Dojo
Plans are Lite $10 a month for 160 credits, Pro $25 for 500 credits, Pro Max $50 for 1,100 credits. Commercial use is included on every plan.
In our test the character still cost 1.5 credits and the voice line 0.5. The talking clips cost 50 credits on OmniHuman, 20 on Kling Avatar and 10 on P-Video Avatar.
Under our refund policy, a first purchase can be refunded within 7 days, in full if under 20% of the credits are used and pro-rated after that.
Hedra questions
What is a good alternative to Hedra?
Pixel Dojo runs OmniHuman, Kling Avatar and P-Video Avatar for talking characters from a still and a voice, plus text to speech, with each price shown before you generate.
How much does Hedra cost?
Hedra lists Basic at $20 a month for 2,000 credits, Pro at $50 for 7,200, Ultra at $100 for 18,000 and Teams at $75 for 14,400. Character 3 is listed from 2.5 cents a second at 540p.
Can I make a cartoon or 3D character talk?
Yes. In our test a 3D style cook and a painted wizard both talked with matching lips on Pixel Dojo avatar models.
Do I need my own voice recording?
No. Write the line and generate it with text to speech, or let P-Video Avatar voice a typed script in the same run.
Which avatar model is the most expressive?
In our test OmniHuman gave the most expressive performance, with a full wave and lively eyebrows. P-Video Avatar was the calmest.
Video Effects New
No prompt? Try a one-click effect
283 ready-made effects, from AI Hug and Earth Zoom In to product spins and listing walkthroughs. Upload a photo and press Generate. No prompt to write.
Try Pixel Dojo as your Hedra alternative
Write a prompt above or start from any example on this page. One subscription covers every model on Pixel Dojo.
Start with OmniHumanSources
Facts about Hedra come from these public pages, checked on Oct 15, 2026. Prices and plans change; check the source before you buy.