Mirelo, a Berlin-based startup, is developing AI technology that adds soundtracks synchronized with video action—addressing a gap in many AI video creation tools which often lack audio support. Earlier this year, Mirelo launched Mirelo SFX v1. 5, an AI model that analyzes videos and adds matching sound effects (SFX). This innovation has caught the attention of venture capitalists preparing for a generative AI boom in gaming. Having operated in stealth mode, Mirelo recently raised a $41 million seed round led by Index Ventures and Andreessen Horowitz, increasing its total funding to $44 million, including earlier investment from Berlin-based Atlantic. The new capital will help the two-year-old company compete against competitors like Sony, Tencent, China’s Kuaishou-owned Kling AI, and ElevenLabs—all of whom have released video-to-SFX models. Mirelo distinguishes itself by focusing more narrowly on sound effect generation but plans to expand its team from 10 employees to possibly 20–30 by the end of next year to boost R&D, product development, and go-to-market efforts. Mirelo’s AI models are available on platforms like Fal. ai and Replicate, with API usage expected to drive most short-term revenue. The startup is also developing Mirelo Studio, a workspace aimed at creators that could eventually support professional-level production.

Mindful of ongoing controversies in generative AI around training data and artists' rights, Mirelo uses public and purchased sound libraries and engages in revenue-sharing partnerships to respect creators. Targeting mostly amateurs and prosumers with a freemium model—featuring a recommended creator plan at €20/month (~$23. 50)—Mirelo aims to “unmute” AI-generated videos, reflecting CEO and co-founder CJ Simon-Gabriel’s belief that audio significantly shapes viewer experience. He quotes George Lucas, stating sound makes up “50% of the movie-going experience, ” if not more, emphasizing how audio transforms the ambiance even with unchanged visuals. Both Simon-Gabriel and co-founder Florian Wenzel are AI researchers and musicians, planning to add AI music generation in the future. However, current demand is stronger for sound effects, partly because the AI sound field is less explored compared to others, offering Mirelo a better opportunity to build a competitive moat. Though the company didn’t disclose its new valuation, Simon-Gabriel revealed it rose substantially from the previous pre-seed round. Their investor base includes influential angels like Mistral CEO Arthur Mensch, Hugging Face’s chief science officer Thomas Wolf, and Fal. ai co-founder Burkay Gur, adding technological credibility and networking advantages. As AI video generators increasingly integrate sound—exemplified by Gemini incorporating DeepMind’s Veo 3. 1 video-to-audio model—Mirelo’s mission feels validated. Simon-Gabriel likens sound’s emergence in AI video to the transition from silent films to “talkies, ” underscoring how critical audio has become in creating immersive experiences.