You Don't Need Eleven Labs Anymore. USE this FREE And OPEN SOURCE Model.

@b0si5

2.9K followers · usually 1.1K views a post

Views

5.4K

Likes

193

Comments

10

Shares

31

SoundTikTok sound · original sound - b0si5

Promoted product

MisoTTS

Miso TTS is an 8 billion, highly emotive text-to-speech model - MisoLabsAI/MisoTTS

Why it worked

The video taps into the desire for free and open-source alternatives to popular paid AI tools, offering a practical solution for users concerned about privacy and cost. The clear demonstration and instructions make it easy for viewers to try the tool themselves.

Summary

The video introduces MisoTTS, a free and open-source AI voice model that serves as an alternative to Eleven Labs. It highlights MisoTTS's ability to clone voices, respond to tone, and run locally on a user's machine, offering a more private and potentially cost-effective solution.

Structure

1Introduction of MisoTTS as a free alternative to Eleven Labs
2Demonstration of MisoTTS's features like voice cloning and tone response
3Explanation of local installation and requirements (Python, GPU)
4Instructions on how to access MisoTTS via GitHub and Google Colab

Original caption

This free AI voice model just dropped and it clones your voice in 10 seconds. No ElevenLabs subscription needed. Here is how to set it up yourself. #bosi #aitools #aivoice #aiagents #claude

More from @b0si5

Product placement

Eleven Labs: ACTS - it is a voice cloning service that the creator is comparing their new model to. Removing it would change the context of the comparison. Miso Labs: ACTS - it is a voice cloning service that the creator is introducing as a free and open-source alternative to Eleven Labs. Removing it would change the entire premise of the video. MisoTTS: ACTS - it is the name of the open-source model being promoted. Removing it would make the video's subject unclear. Google Colab: ACTS - it is a platform used to run the MisoTTS model. Removing it would make the setup instructions incomplete.

Call to action

Go to MisoLabs.ai for a live demo, or head over to GitHub to clone their repo and use Google Colab to generate clips.

On-screen text

You Don't Need Eleven Labs Anymore. USE this FREE And OPEN SOURCE Model.

Transcript

This just came out yesterday and I thought, you know what, let me make a video on it. I was scrolling through X and it came across something called Miso one. It is an AI voice model, completely free and open source. Personally, I use Eleven Labs, right? But after seeing what this thing can do and how realistic it sounds, for me, it could actually replace Eleven Labs. The ones that can actually respond to your tone when you speak to it. So if you whisper to it, it's gonna whisper back, right? It's crazy, right? You can clone your voice as well. Well, it runs locally on your machine, so your audio never actually goes anywhere. So to try it right now, you can go to MisoLabs.ai. There's actually a live demo up, but you know, if you're a bit more technical, you can run it locally. It will need Python 3.1 and a GPU with enough VRAM to run this model and generate audio for yourself. Head over to their GitHub, clone their repo and then use Google Colab to basically generate clips from this model. Right? I'm not sure some of you may be thinking, what is this guy talking about? Um, that's fine. Their API is not actually out yet. So you can wait for their API to be released, and then that'll be an easier way of using it. But honestly, for anything voiceovers, content, voice AI agents, so Miso one might just be the one. Try it out honestly. Let me know how it goes. Follow for more free game.