At the TechCrunch Disrupt 2025 conference, ElevenLabs co-founder and CEO Mati Staniszewski shared his perspective on the future of AI audio, predicting that AI models in the space will eventually become commoditized. Despite this long-term outlook, he emphasized that building proprietary models remains a critical advantage in the short term. Speaking on stage, Staniszewski acknowledged that ElevenLabs has made significant progress in overcoming key model architecture challenges, a focus that will continue for the next 12 to 24 months. “Over the long term, it will commoditize — over the next couple of years,” he said. “Even if there are differences — which I think will be the truth for some voices, some languages — on their own, the differences will be smaller.” When asked why ElevenLabs would continue investing in model development if it expects commoditization, Staniszewski explained that today, high-quality audio generation is still a major differentiator. “If the AI voices or interactions don’t sound good, that’s still a problem that needs to be solved,” he said. “The only way to solve it is by building the models yourself. And over time, other players will do the same.” He also noted that while the core audio models may converge, there will still be value in using different models for different applications, especially in complex or specialized use cases. However, the next phase of development will likely involve more multi-modal and integrated approaches. “You’ll create audio and video at the same time, or audio and large language models in a conversational setting,” he said, citing Google’s Veo 3 as an example of what’s possible when models are combined. Looking ahead, ElevenLabs plans to expand its strategy by forming partnerships with other companies and leveraging open-source technologies. The goal is to combine its deep audio expertise with the capabilities of other models to create more powerful, seamless experiences. Ultimately, Staniszewski said the company’s vision is to focus on both model innovation and real-world applications. “The same way software and hardware were the magic for Apple, we think the product and AI will be the magic for the best use cases in this new generation,” he added.