Stable Audio 3 (https://huggingface.co/collections/stabilityai/stable-audio-3) is the new version of the music generation model of the Stable Audio family already seen here.
Open source for 3 of its 4 variants, generates instrumental music tracks and sound effects from an instruction, with sounds/music that can be added as cues.

Generations up to 6 min 20 from two possible inputs;

• since https://stableaudio.com/, with account
• since https://huggingface.co/spaces/stabilityai/stable-audio-3) gives access to the 3 models opened without an account.

Trained on nearly 1.3 million fully licensed recordings with published attribution (see http://info.stability.ai/attributions).

Export wav from HuggingFace and mp3 or mp4 video from stableaudio.com

On stableaudio.com, 6 generations per month with Stable Audio 3 in free version.
HuggingFace Space is accessible without an account, with the usage quotas linked to HuggingFace.

Sample audio for testing: https://commons.wikimedia.org/wiki/File:Audio_files_of_spanish_guitar_chords_28.wav