Text or audio in. A talking video out. One call.
The dialogue endpoint chains speech and lip-sync for you: send a face and a script, or a face and an audio file. 12 lip-sync and avatar models, at base model cost plus a flat 20% platform fee, shown before you spend.
- 12 lip-sync and avatar models
- Text or audio input
- Quote before you run
- No plan
A talking video is usually two vendors and a glue script.
- A text-to-speech vendor for the voice
- A separate avatar or lip-sync vendor for the face
- Your own code to pass audio from one to the other, and two bills
- Avatar tools sold on monthly plans priced in minutes or credits
- One call: text, voice and face in, video out
- Or bring your own audio file and skip the speech step
- One quote covers both legs before anything runs
- Speech, lip-sync and everything else on one balance
What you can build
A presenter that reads your script in any of 168 library voices.
Dub the audio, then lip-sync the face to the new language.
Multiple speakers in one request for short scenes.
Pass an audio URL and the model lip-syncs the face to it.
Re-sync a video to new audio. Kling LipSync is $0.0168/s.
One request, both legs
estimate: true returns the price of the speech and the lip-sync together. Remove it to run.
curl -X POST https://api.fairstack.ai/v1/dialogue/generate \
-H "Authorization: Bearer $FAIRSTACK_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"text": "Welcome back. Here is what changed this week.",
"character_image_url": "https://example.com/presenter.png",
"talking_head_model": "infinitetalk",
"resolution": "480p",
"estimate": true
}'Lip-sync and avatar prices
Live catalog prices. Per-second models are billed on the length of the output.
Live catalog prices: base model cost plus a flat 20% platform fee. Every call is priced before it runs. All models
How it runs
- 01Quote
Send the request with estimate set to true. You get the speech cost, the lip-sync cost and the total.
- 02Run
Drop estimate. The call answers in milliseconds with a job id and runs both legs in the background.
- 03Collect
Poll the job or take a webhook. You get the video URL and the final cost of each leg.
Frequently asked questions
Can I send my own audio instead of text? +
Yes. Pass audio_url and the speech step is skipped. The face is lip-synced to your audio.
What does a talking video cost? +
It depends on the model and the length. InfiniteTalk is $0.30 per generation (480p). A text request also pays for the speech it generates. Every price is base model cost plus a flat 20% platform fee, shown before you spend.
Is it synchronous? +
No. The call validates your input and answers with a job id straight away, then runs in the background. Poll the job, or pass wait to block until the video is ready.
Do I need a plan? +
No. FairStack has no plans. Add funds to one balance, $1 minimum, and pay per generation. Your balance never expires.
Still have questions? We're here to help.