Lyria 2 is a text-to-music model that takes a written description and returns an original audio track in stereo at 48kHz. It removes the barrier between having an idea for a sound and actually hearing it, without needing any recording equipment, software instruments, or music production background. The model reads both a main prompt and an optional negative prompt, giving you two levers to shape the output. You describe what you want, like a slow jazz piano piece or an upbeat electronic loop, and use the negative prompt to push specific elements out of the result, such as distorted guitar or background noise. Output comes back as broadcast-quality stereo audio, ready to use without additional processing. Lyria 2 fits naturally into content workflows where original music is needed but licensing costs or production time are constraints. A seed parameter lets you reproduce a result exactly, which is useful when you find a sound you like and want to iterate on it. Open the model, describe what you need, and have a usable audio file in seconds.
Lyria 2 is a music generation model that converts text descriptions into original stereo audio at 48kHz quality, available on Picasso IA. It addresses a real constraint for content creators who need original music but have no recording setup or licensing budget. A video editor, for instance, can describe the exact mood and instrumentation they want and receive a finished track in seconds. The negative prompt feature adds a second layer of control, letting you steer the result away from specific elements you want to avoid. No prior music knowledge is needed, and the output is clean, broadcast-ready audio.
Do I need programming skills or technical knowledge to use this? No, just open Lyria 2 on Picasso IA, adjust the settings you want, and hit generate.
Is it free to try? Yes, you can run Lyria 2 on Picasso IA without a paid subscription to test the results. A few free runs let you hear what the model produces before committing.
How long does it take to get results? Most generations return a finished audio file within a few seconds. Prompt complexity has minimal impact on generation time.
What output format does Lyria 2 produce? Lyria 2 outputs stereo audio at 48kHz, which is compatible with standard video editors, DAWs, and game engines without any conversion.
Can I control what the music sounds like? Yes. The text prompt shapes the genre, mood, and instrumentation, while the negative prompt lets you exclude specific elements like vocals or drums. Running the same prompt multiple times with different seeds also gives you distinct variations to choose from.
How many times can I run the model? You can run it as many times as you want. Each run with a different seed or prompt gives you a distinct result.
Where can I use the audio it generates? The audio files are yours to use in personal projects, videos, games, and other creative work. There are no watermarks on the output.
Everything this model can do for you
Delivers broadcast-quality audio that drops into video timelines and audio editors without resampling.
Converts a written description of genre, instruments, and mood directly into a finished audio track.
Lets you specify what to exclude from the output, such as drums or vocals, for cleaner results.
Reuse the same seed to get an identical track across multiple runs, useful for iterating on variations.
Produces original music from text alone, no recording equipment or music theory background needed.
Generates tracks spanning ambient, cinematic, electronic, acoustic, and more from a single prompt.
Returns a finished audio file in seconds, so you can run multiple variations without long waits.
Futuristic country music, steel guitar, huge 808s, synth wave elements space western cosmic twang soaring vocals
A cinematic orchestral piece with epic brass swells, thunderous timpani rolls and soaring string melodies
