The post Google DeepMind Lyria 3 Launches AI Music Generation in Gemini App appeared on BitcoinEthereumNews.com.

Joerg Hiller
Feb 23, 2026 17:52

Google’s Lyria 3 AI music model now creates 30-second tracks from text or images in Gemini app, featuring automatic lyrics and SynthID watermarking.

Google DeepMind has rolled out Lyria 3, its most advanced generative music model, within the Gemini app as of February 18, 2026. The tool lets users create original 30-second tracks using text prompts, photos, or video uploads—a notable expansion of AI capabilities into creative audio production. The upgrade from Lyria 2 brings automatic lyrics generation, meaning users no longer need to write their own words. Simply describe a theme like “a love song” or “a track about success,” and the model handles the rest. For those who prefer control, custom lyrics still work—just prefix them with “Lyrics:” in your prompt. How It Works Access Lyria 3 through the “create music” option in the Gemini app. From there, you’ve got several input methods. Text prompts can specify genre, era, instruments, vocal style, and song dynamics. Want ’90s skate punk with fast drums? Done. Prefer blending K-pop with Motown influences? The model accommodates genre mashups. Visual inputs add another dimension. Upload a vacation photo, pet picture, or original artwork, and Lyria analyzes elements like subjects, clothing, and backgrounds to generate a musical match. The same applies to video clips. Creative Controls Specificity pays off with Lyria 3. Users can define whether vocals come from a male or female singer, specify vocal range from baritone to soprano, and add texture descriptors like “gravelly” or “breathy.” Instrumental choices extend beyond genre defaults—adding ’80s synth to a 1950s jazz foundation is fair game. Song dynamics also respond to prompting. A quiet piano intro building to an explosive chorus, purely instrumental sections, or vocals that split into harmonies as…

---

Filed under: News - @ February 24, 2026 8:25 am