Need AI Face Swap and Voice Cloning on the Cheap!

Hubba

Elite Member
Joined
Feb 6, 2010
Messages
2,023
Reaction score
1,099
I want to make a series of product review and promotional videos, BUT I don't want my real face and voice "out there." I've looked at https://magicam.ai/, but it only does "real-time" streaming face swapping and voice cloning. Plus, my machine doesn't have a strong enough gpu.

Here's what I'm looking for:
1) Ability to upload 5 - 10 minute video and have the face and voice swapped out.
2) Need ability to have the same voice and face for future videos.
3) CHEAP - There are many services that do what I'm asking, BUT most won't do over 5 minutes and they are EXPENSIVE!

If anyone has a good solution let me know. If need be, I'm willing to rent gpu server for this.

Thanks!
 
higgsfield ai is easiest and cheapest on market to use

Ive been using it for a couple months its amazing
Unlimited nano banana pro is amazing if you need static ads
 
higgsfield ai is easiest and cheapest on market to use

Ive been using it for a couple months its amazing
Unlimited nano banana pro is amazing if you need static ads
Can you explain the models and pricing for what I'm wanting to do? I was looking at their plans and for the Pro plan they list the following:

Higgsfield DoP Standard 720p 7 credits/3s - 200 videos
Higgsfield DoP Turbo 720p 5 credits/3s - 85 videos
Sora 2 720p 10 credits/4s - 60 videos

They don't explain how long a "video" is. If I want to face and voice swap my uploaded videos in 720p how many credits would it take for a 5 minute video?
I'm confused about credits vs videos, etc.

Thanks!
 
What about Replicate?
I am using it mostly for images, but there is nice wide selection of Video models too. No monthly fees, just pay as you go
https://replicate.com/collections/text-to-video
 
You might want to check out DeepBrain or Synthesia, which offer AI-based face and voice swapping for videos. They allow you to upload content and replace faces/voices while maintaining consistency across multiple videos. While these services can be costly, you can also look into RunwayML or D-ID for more affordable options that offer video editing tools with AI features. Renting a GPU server for OpenAI's DALL·E or similar tools might also be a cost-effective solution if you're comfortable with some technical setup.
 
UDPATE: Regarding higgsfield. I contacted them. They only support 5 - 10 second videos, so I would have to stitch "segments" together according to support. For $18/month there would only be enough credits to do 1 - 5 minute video at the lowest resolution. That won't work at all.
 
UDPATE: Regarding higgsfield. I contacted them. They only support 5 - 10 second videos, so I would have to stitch "segments" together according to support. For $18/month there would only be enough credits to do 1 - 5 minute video at the lowest resolution. That won't work at all.
For quality higgsfield might be your best bet. The videos are going to take effort either way to produce so it does suck that you have to stitch them together but they are unbeated for video generation and budget

Ive tried other platforms and they are really bad
 
I want to make a series of product review and promotional videos, BUT I don't want my real face and voice "out there." I've looked at https://magicam.ai/, but it only does "real-time" streaming face swapping and voice cloning. Plus, my machine doesn't have a strong enough gpu.

Here's what I'm looking for:
1) Ability to upload 5 - 10 minute video and have the face and voice swapped out.
2) Need ability to have the same voice and face for future videos.
3) CHEAP - There are many services that do what I'm asking, BUT most won't do over 5 minutes and they are EXPENSIVE!

If anyone has a good solution let me know. If need be, I'm willing to rent gpu server for this.

Thanks!
You can also use Google’s voice features to easily add voice-overs to your videos, this is a method many people still use.
 
UDPATE: Regarding higgsfield. I contacted them. They only support 5 - 10 second videos, so I would have to stitch "segments" together according to support. For $18/month there would only be enough credits to do 1 - 5 minute video at the lowest resolution. That won't work at all.
I'm working on a video with a face (not swap) , reduced the presence by cutting to broll and mixing with a separate audio track from the start.
 
I want to make a series of product review and promotional videos, BUT I don't want my real face and voice "out there." I've looked at https://magicam.ai/, but it only does "real-time" streaming face swapping and voice cloning. Plus, my machine doesn't have a strong enough gpu.

Here's what I'm looking for:
1) Ability to upload 5 - 10 minute video and have the face and voice swapped out.
2) Need ability to have the same voice and face for future videos.
3) CHEAP - There are many services that do what I'm asking, BUT most won't do over 5 minutes and they are EXPENSIVE!

If anyone has a good solution let me know. If need be, I'm willing to rent gpu server for this.

Thanks!
Try Wan 2.2 from Alibaba. It is self hosted and open source. Meaning you need to make use of your local resources to generate the video and face swap. Can't ask for more since it is free. If you know coding, you might be able to modify it to loosen any restrictions imposed. If you have a power computer sitting at home, this would be idea as you will be saving alot of money.
 
Try Wan 2.2 from Alibaba. It is self hosted and open source. Meaning you need to make use of your local resources to generate the video and face swap. Can't ask for more since it is free. If you know coding, you might be able to modify it to loosen any restrictions imposed. If you have a power computer sitting at home, this would be idea as you will be saving alot of money.
That's the problem. I don't have a decent gpu, and even used computers and gpus are crazy expensive now. Even a $1,000 used computer takes FOREVER for AI image generation, let alone proper AI video. Can Wan 2.2 be installed on a rented gpu server? That might be the way to go.
 
That's the problem. I don't have a decent gpu, and even used computers and gpus are crazy expensive now. Even a $1,000 used computer takes FOREVER for AI image generation, let alone proper AI video. Can Wan 2.2 be installed on a rented gpu server? That might be the way to go.
You can try the below. Alternatively, create multi stealth account to leverage on free tokens for new accounts from Alibaba GPU for Wan 2.2. Remember to cover your tracks when using this method and don't over abuse.
  • RunPod Serverless: Offers specific pricing for Wan 2.2 models:
    • Wan 2.2 T2V/I2V (720p): ~$0.30 per 5-second request.
    • Wan 2.6 T2V/I2V: ~$0.50 per 5-second request.
 
Absolutely. The rankings come as the result, but they are not the base. High search intent, helpful content, topical authority, trust signals, and good user experience will ensure SEO success instead of temporary rankings gain.
 
The best models like flux, seedance and comfy ui workflows you use a runpod gpu charged by the hour when working. Huggingface is a wrapper saas.
 
I want to make a series of product review and promotional videos, BUT I don't want my real face and voice "out there." I've looked at https://magicam.ai/, but it only does "real-time" streaming face swapping and voice cloning. Plus, my machine doesn't have a strong enough gpu.

Here's what I'm looking for:
1) Ability to upload 5 - 10 minute video and have the face and voice swapped out.
2) Need ability to have the same voice and face for future videos.
3) CHEAP - There are many services that do what I'm asking, BUT most won't do over 5 minutes and they are EXPENSIVE!

If anyone has a good solution let me know. If need be, I'm willing to rent gpu server for this.

Thanks!
If you don’t have a GPU at all, RunPod nd Wan/ComfyUI pipeline appear to be the most reasonable escape route available. the pay-per-use option is superior to being trapped in a costly subscription system.
 
Back
Top