DeepInfra adds serverless ACE-Step 1.5 XL
DeepInfra has added serverless inference API support for the open-source text-to-music model ACE-Step 1.5 XL. The integration runs the planning step by default to optimize musicality, enabling developers to generate high-fidelity, full-length songs with vocals and instrumentation from text prompts.
Hosting ACE-Step 1.5 XL on DeepInfra brings open-source music generation closer to production readiness, offering a direct threat to closed platforms like Suno and Udio by reducing inference friction and cost.
- –**Open-source parity**: ACE-Step 1.5 XL represents a significant step forward in open-source audio, matching the multi-instrument and vocal capabilities previously locked behind proprietary APIs.
- –**Developer accessibility**: DeepInfra's serverless hosting model allows developers to easily integrate high-quality music generation into applications without managing complex GPU infrastructure.
- –**Default planning step**: Running the planning step by default ensures the generated compositions maintain structural and musical coherence, addressing a common weakness in raw text-to-audio outputs.
DISCOVERED
50d ago
2026-06-12
PUBLISHED
50d ago
2026-06-12
RELEVANCE
AUTHOR
DeepInfra