Lyria 3

Lyria 3 is an online AI music generator that creates complete, royalty‑free songs from textual descriptions or visual inputs. Users provide a prompt specifying genre, mood, tempo, and desired lyrics, or upload a photo or video, and the system produces a studio‑quality stereo track with instrumental accompaniment, custom lyrics, and synthesized vocals. The output is delivered in 48 kHz high‑fidelity audio, ready for use in media such as videos, podcasts, games, or advertisements.

The service targets creators who lack musical training or studio equipment but need original music quickly, including content producers, advertisers, and hobbyist musicians. It supports multiple vocal languages and a range of styles, from pop and hip‑hop to cinematic orchestral and electronic EDM. The underlying model is a third‑generation AI music system built on latent diffusion architecture and trained on extensive licensed datasets, enabling autonomous songwriting and multimodal generation.

All generated tracks are claimed to be 100 % royalty‑free, allowing users to download and use the audio without additional licensing. The platform offers a web interface for prompt entry, preview, and download, and provides free credits upon sign‑in for initial experimentation.