cd /news/generative-ai/google-releases-lyria-3-5-music-gene… · home topics generative-ai article
[ARTICLE · art-122970] src=mlq.ai ↗ pub= topic=generative-ai verified=true sentiment=· neutral

Google releases Lyria 3.5 music generation in Gemini and its API

Google released its Lyria 3.5 music-generation model in the Gemini app and Gemini API on September 4, enabling full-length songs from text or image prompts with 44.1 kHz stereo output and controllable structures. The API is priced at $0.08 per generated song with no free tier, and safety filters block prompts for specific artist voices or copyrighted lyrics while every track carries a SynthID watermark.

read2 min views2 publishedSep 8, 2026
Google releases Lyria 3.5 music generation in Gemini and its API
Image: Mlq (auto-discovered)
  • Lyria 3.5 generates full-length songs from text or images, with 44.1 kHz stereo output and controllable structures such as verses, choruses and bridges. <sup>[1]</sup>
  • Google lists API pricing at $0.08 per full song, with no free tier for Lyria 3.5. <sup>[2]</sup>
  • Safety filters block prompts seeking specific artist voices or copyrighted lyrics, and every generated track carries an inaudible SynthID watermark. <sup>[3]</sup>

Google released its Lyria 3.5 music-generation model in the Gemini app and Gemini API on September 4, extending the system beyond the 30-second clips offered by the earlier Lyria 3 rollout. Google says Lyria 3.5 is available globally on Gemini’s web and mobile apps, as well as in Google AI Studio, Flow Music and Google Vids. [4]

Full songs, paid API access #

Lyria 3.5 can generate songs lasting a couple of minutes from text or image prompts, including vocals, timed lyrics and full instrumental arrangements. The Gemini API documentation lists the model at $0.08 per generated song and says no free tier is available. [1][2]

Google has not disclosed consumer access tiers, daily limits or regional exceptions in the launch post. Its announcement describes the Gemini rollout as global, while the developer documentation identifies Lyria 3.5 as a preview model. [4][5]

Licensing claims remain broad #

Google’s public materials do not identify a specific licensed training catalog or name major music-rights organizations for Lyria 3.5. The company says it has worked with artists on safeguards and that YouTube’s Artist Partnerships team supported its music-industry partnerships. Google has separately said Flow Music is working with Believe and TuneCore artists, producers and songwriters, and that its training materials are drawn from content YouTube and Google have rights to use under agreements, terms of service and applicable law. [6][7]

The API blocks prompts requesting a specific artist’s voice or copyrighted lyrics. Generated audio is embedded with SynthID, while Google says its datasets use filtering and labeling to reduce harmful lyrics. Lyria 3.5 currently generates music in a single turn rather than supporting iterative edits of an existing track. [3]

Companies mentioned #

Further sources #

The stories that matter, in one email. Free — unsubscribe anytime.

── more in #generative-ai 4 stories · sorted by recency
── more on @google 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
Live at https://your-agent.zahid.host
Get free account → Pricing
from €0/mo · no card required
LIVE [news/google-releases-lyri…] indexed:0 read:2min 2026-09-08 ·