arxiv.org/abs/2608.04142
TL;DR - FAD without a background set.
Checkout our paper, I think that the idea is pretty neat and practical!
A joint project with the amazing Harel Pogoda and @adiyossLC .
See us at ISMIR 2026 🎻🎻🎻
Alon Ziv retweeted
Expecting A Minor update? We prefer to B Major. 🎶
@GoogleDeepMind just released Lyria 3.5, and we’ve plugged it directly into @GoogleFlowMusic to power your next session.
What this means for your tracks:
✨ Better prompt adherence: Flow Music understands your creative direction with much more accuracy. Set exact BPMs and export stems for full-length songs!
🎤 Upgraded vocals: More expressive, dynamic, and rich singing.
🎸 Richer musicality: Arrangements flow more naturally from note to note.
Hear it for yourself and start creating at flowmusic.google 👇
Alon Ziv retweeted
Meet Lyria 3.5, the newest music model from @GoogleDeepMind, now powering Flow Music:
🎤 Vocals: More expressive, dynamic singing
🎸 Musicality: Richer arrangements with tracks that flow naturally.
🎛️ Controls: Better at following creative directions - now featuring BPM control and exportable stems for full-length songs!
Try it now at flowmusic.google! 🧵
Alon Ziv retweeted
Introducing Magenta RealTime 2 (MRT2): the live music model you can play as an instrument.
MRT2 offers MIDI and prompt controls, and runs natively on a MacBook with <200ms latency.
Open weights. Open source inference engine. Suite of apps and plugins.
Hear what it can do and try it out for yourself below 🧵
Alon Ziv retweeted
We’ve released TECCI, a new dataset focusing on challenging image edits! (Including visual puzzles and much more)
Like DOCCI, all images are donated to the public domain, so you get the bytes with the data. See Aishwarya’s thread for more details and pointers.
Releasing a new image editing benchmark -- TECCI: Tricky Edits of Collected and Curated Images.
Paper: arxiv.org/abs/2606.01213
Project website: google-deepmind.github.io/te…
arxiv:
arxiv.org/abs/2512.10264
demo page:
lonzi.github.io/mr_flowdpo_d…
code:
github.com/lonzi/mrflow_dpo
Alon Ziv retweeted
1/ A new era, a new name, and a new feature drop. 🚀
ProducerAI is officially "Google Flow Music"! We're still the same team, with the same mission: building tools that push the boundaries of human creativity.
To kick off this new chapter, we're launching new capabilities to give you more control over your tracks.
Let's dive in: 🧵👇
Alon Ziv retweeted
Today, we're open sourcing the code behind "The Infinite Crate," our VST/AU plugin that lets you play with Lyria RealTime directly inside your favorite DAW! 🎧🎶
Since we released The Infinite Crate last year, it’s been used by some of our favorite artists - including a wonderful showcase with @daitomanabe in Tokyo - and being featured as a top new music tool at NAMM 2026.
We’re now fully open sourcing the plugin for developers to fork, modify, and make their own under the permissive Apache 2.0 license.
💾 Get it here: github.com/magenta/the-infin…
Alon Ziv retweeted
Alon Ziv, Sanyuan Chen, Andros Tjandra, Yossi Adi, Wei-Ning Hsu, Bowen Shi, "MR-FlowDPO: Multi-Reward Direct Preference Optimization for Flow-Matching Text-to-Music Generation," arxiv.org/abs/2512.10264
Alon Ziv retweeted
Introducing Spaces 🪐
Artists master hundreds of tools in their craft. Our vision is for artists to not just USE these tools, but to CREATE any instruments, plugins, and experiences they imagine.
That vision is coming to life today on producer.ai
can i get
myself, @lonziks, becoming a goldfish that plays the drums under the sea?
@producer_ai
Alon Ziv retweeted
Feedback from the SF @exploratorium staff after making music with countless school groups using @producer_ai 🙌
“It opens up for them ideas of AI’s power and also creativity.
It helps people understand that AI isn’t monolithic
It opens creativity not only with AI, but with collaboration with other visitors.”
Truly excited from this one!
We started vibes as a rather experimental research project and ended up with an intuitive yet abstract style conditioning for producer.ai.
People make good stuff with it! see playlist in the thread 🎧
This quoted post is unavailable.
producer.ai/song/2c46da8f-aa… -
a jazzy one, pretty wild IMO, from our Fuzz 2.0 Raw model
Alon Ziv retweeted
"It allows you to completely break your patterns as a producer." - @harpt
We were recently joined by Tushar Apte, a professional songwriter and producer based in LA, to hear how tools like Producer allow him to find new inspiration and break out of his creative patterns.
A vibe is a 10 second audio prompt from which the model takes motifs.
It could be a melodic pattern, a rhythmic pattern, the production style, or the characteristics of the vocalist.
Mixing vibes takes u to the uncharted territory of motif blending
Alon Ziv retweeted
“You can create bad AI music just like you can create bad regularly recorded music.”
Max Vehuni from @slenderbodies shares his perspective, as a professional musician, on AI music during his Coffee Table Session at Producer HQ.











