Dance like nobody’s watching. Let them watch. Let them love it. | Head of Audio Research @StabilityAI | 🎵: https://teapaw.org/t.co/WzwP0l9L6n
Bellevue, WA
Joined June 2008
- Tweets4.2K
- Following580
- Followers1.5K
- Likes9.4K
zach 🏔🎶 retweeted
Stop pulling from the same sample library as everyone else. Build one that's entirely your own.
We're hosting a deep dive into the Stable Audio plugin: prompt-based sound generation that runs directly inside your DAW.
📅 Wed, Sept 30 @ 11am ET
🔗 youtube.com/live/zzKUjjhVZbs
zach 🏔🎶 retweeted
You know the moment: mid-session, losing time scrolling a sample library.
Sept. 30th, we're skipping the library. Generate samples in key and in tempo with the Stable Audio plugin, then process them into something original.
📅 Wed, Sept 30 @ 11am ET
🔗 youtube.com/live/zzKUjjhVZbs
zach 🏔🎶 retweeted
Most drum machines can't keep up with a live performance.
Team StableGroove fixed that at the Montreal @MusicHackspace Hackathon: a real-time AI drum machine on Stable Audio 3.0, generating fresh samples locally without breaking the rhythm.
Watch Nav Kumar break it down.
zach 🏔🎶 retweeted
“open source local models are important b/c musicians should be using instruments not services”
zach 🏔🎶 retweeted
Music hackers building on Stable Audio 3.0 at the Montreal Music Technology Hackathon told us why open weight models matter to them: transparency, artistic control, and an actual say in how the technology gets used.
Watch what they said 👇
zach 🏔🎶 retweeted
Fresh results from July 2026 at Music Arena.
Happy to see the community is finding Stable Audio 3 Medium as useful as we do.
It has been independently evaluated, and Medium is the second-best and fastest model for instrumental music generation, with open weights!
zach 🏔🎶 retweeted
Most AI audio models have never heard a maqam.
Team Motif fine-tuned Stable Audio 3.0 on Arabic maqam, built an Ableton plugin for microtonal style transfer, and won our Stable Audio 3.0 Challenge at Music Hackspace running locally on device.
Watch Jad Al Masri break it down 👇
zach 🏔🎶 retweeted
Underfit: Train Stable Audio 3 Loras on your Local PC
Underfit (by @dadabots) lets you train your own Stable Audio 3 Loras.
Feed your own music samples to train a Lora, and use that as input to generate songs.
1-click launcher (linux + windows NVIDIA GPU required)
The community is already doing awesome things with Stable Audio 3! Check it out!
Check out Awesome Stable Audio, a curated list of integrations and extensions for stable audio. If I missed anything or you build something cool, please submit a PR and I will add you!
github.com/Stability-AI/Awes…
zach 🏔🎶 retweeted
i see ppl making chord generators with the
ableton extensions sdk
im making some everything generators with it
first up: stable audio 3
next up ace-step-1.5
zach 🏔🎶 retweeted
Fast workflow for Stable Audio 3 and Ableton Live 12 on M4 mac!
Instructions here:
github.com/Stability-AI/stab…
Make sure you have already downloaded github.com/Stability-AI/stab…
zach 🏔🎶 retweeted
1/7
recently i started exploring mechinterp, and i started with audio diffusion.
i spent some time cracking open stable audio 3 to see if mood has a zip code in its residual stream.
it does! it's layer 11 and you can dial it. (audio results below) 🧵
zach 🏔🎶 retweeted
Made a club anthem entirely out of the Discord notification sound. Stable audio inpainting.
zach 🏔🎶 retweeted
Ringtone salsa --- generated with stable audio inpainting
zach 🏔🎶 retweeted
Run Stable Audio 3 on ANY Computer with NO VRAM
1-click launcher for the official Stable Audio 3 gradio app.
1. Cross Platform: Works on ALL OS (Mac, Linux, Windows) out of the box
2. No GPU Needed: Even runs on CPU.
3. Ultra Fast: Generate a 2 minute clip in split seconds
Meet Stable Audio 3.0, the open-weight model family built for artistic experimentation.
This is our open invitation to experiment with generative audio. We believe the best innovations are still waiting to be built.
The 4-1-1 on 3.0:
📣 You own your outputs, and can distribute and commercialize them under the Stability AI Community License (up to $1 million in revenue).
🎵 New and improved capabilities include variable-length generation up to six minutes, and full song composition on portable devices, no GPU required.
✅ Trained on a fully licensed dataset.
🎨 You can customize the models on your own library with support for LoRa training, which we’ve documented for the first time.
More on the models 👇
