ElevenLabs launches Music v2 with mid-track genre switching and section-by-section composition
ElevenLabs released Music v2, an AI music generation model that can switch genres within a single track and build songs section-by-section. The model, trained on licensed data cleared for commercial use, can transition from opera to heavy metal, handle fast rap, and add sound effects while maintaining coherence.
ElevenLabs launches Music v2 with mid-track genre switching and section-by-section composition
ElevenLabs released Music v2, an AI music generation model that can switch genres within a single track, 10 months after launching its first music generation model.
Key capabilities
According to ElevenLabs, Music v2 can:
- Switch between genres mid-track, transitioning from opera to heavy metal and back
- Handle fast rap while maintaining coherence
- Add non-musical sound effects to tracks
- Edit specific sections of a song using prompts without affecting other parts
- Build songs section-by-section (intro, verse, chorus) and stitch them together
- Generate vocals and complex compositions across multiple languages
The model marks a shift from generating short clips to constructing full songs with discrete sections that can be assembled.
Licensing and commercial use
ElevenLabs emphasized that Music v2 is trained on licensed data and cleared for commercial use. This approach differs from competitors Suno and Udio, which both face ongoing copyright litigation from major labels.
Availability
Music v2 is available through:
- ElevenCreative tool for marketing and branding teams
- ElevenMusic platform for AI-generated song creation
- ElevenAPI (coming soon)
Pricing details were not disclosed.
Market context
The release intensifies competition in AI music generation. In recent months:
- Google added song covers, section editing, and music video generation to its Flow Music tool at Google I/O
- Stability AI released new music generation capabilities
- Suno launched updated models for longer, more complex tracks
What this means
Music v2's section-based composition approach addresses a key limitation in AI music generation: the inability to make targeted edits. By allowing artists to modify specific parts without regenerating entire tracks, ElevenLabs moves closer to professional music production workflows. The emphasis on licensed training data positions the company to avoid the legal challenges facing competitors, though questions remain about whether labels will embrace AI-generated music at scale. The mid-track genre switching capability, while technically impressive, may have limited practical applications beyond novelty tracks and experimental compositions.
Related Articles
Liquid AI releases open d1-3B decision model: 16 ms on Jetson AGX Thor, 48.57 on Decision Index
Liquid AI released two open-weight decision models, d1-3B (text and image) and the experimental d1-omni-600M (text with image or audio). Unlike generative models, they answer in a single forward pass, and Liquid AI claims d1-3B scores 48.57 on its Decision Index 0.2.1, ahead of all 4B and 9B models it tested.
Google releases EmbeddingGemma 2: 740M-parameter multimodal embedding model under Apache 2.0
Google announced EmbeddingGemma 2, a 740M-parameter natively multimodal embedding model built on the Gemma 4 architecture and released under Apache 2.0. Google says the quantized model needs about 191MB of active RAM for text-only weights and about 567MB for the full multimodal model on a Pixel 11 Pro. Google also launched a Mac app, AI Edge Foresight, to demonstrate it.
Google releases Nano Banana 2.1 image model: $1.50/$30 per 1M tokens, 66K context
Google's Nano Banana 2.1 (Gemini Nano Banana 2.1) is an image generation and editing model on the Flash tier, listed on OpenRouter at $1.50 input and $30 output per 1M tokens with a 66K context window. It supports 1K, 2K, and 4K output and succeeds Nano Banana 2 and Nano Banana Pro, according to the listing.
Mistral Large 4 enters public preview: 1T-parameter open-weight multimodal model, weights due by end of October
Mistral AI has launched a public preview of Mistral Large 4, a 1-trillion-parameter natively multimodal model with 49 billion active parameters. The preview API is live on Mistral Studio, and open weights are promised by the end of October 2026. Pricing and context window have not been disclosed.
Comments
Loading...