Mirelo SFX V2.0 lands on fal

Mirelo is targeting one of generative video’s biggest remaining gaps: production-ready sound that understands edits, speech, and scene context.

• –Video-to-audio workflows can add effects without manually rebuilding the soundtrack
• –Dialogue preservation makes the model more useful for narrative footage and creator content
• –Continuity across cuts should reduce jarring changes in ambience and Foley
• –Text-to-audio support expands Mirelo beyond automatic video sound design
• –fal availability gives developers an accessible API path for integrating generated audio into pipelines