11:21
11:00
21:05
20:16
14:35
12:00
11:21
11:00
21:05
20:16
14:35
12:00
11:21
11:00
21:05
20:16
14:35
12:00
11:21
11:00
21:05
20:16
14:35
12:00
Google has introduced a music creation section in Gemini powered by its Lyria 3 model, which is currently in beta.
Users can generate 30-second audio clips from text prompts or upload an image as a reference. The model supports eight languages: English, German, Spanish, French, Hindi, Japanese, Korean, and Portuguese. The service also automatically creates cover art using Nano Banana.

All tracks generated by Lyria 3 are tagged with the SynthID digital watermark. Google has also added an AI music detector in Gemini, allowing users to upload and verify whether a track was AI-generated.
Lyria 3 is now also available to all YouTube users via the Dream Track feature, after previously being limited to testing in the United States.

