ElevenLabs CEO Mati Staniszewski, speaking at the TechCrunch Disrupt 2025 conference, said that AI voice models will eventually become commoditized — meaning they will become widespread and standardized products. According to him, within the next few years, there will be little difference between these technologies, as everyone will reach a similar level of quality.
But if models are destined to become standardized, why does ElevenLabs continue to focus on creating them? Staniszewski explained that, in the short term, having the best-performing model still represents a major competitive advantage. He noted that the only way to solve issues such as AI voices not sounding fully natural is to continue developing and training their own models.
In conclusion, Staniszewski predicted that future AI models will be multimodal — capable of processing voice, video, and text simultaneously. He compared the company’s strategy to Apple’s, saying:
“What made Apple magical was the union of software and hardware. The magic of our generation will be the union of AI and the finished product.”
