Inworld AI Launches TTS-1.5, a Groundbreaking Voice AI model for the Next Wave of Realtime AI Applications

MOUNTAIN VIEW, Calif., Jan. 21, 2026 (GLOBE NEWSWIRE) -- January 21, 2026 - Inworld AI, a research lab developing production-grade AI models and infrastructure for the next wave of AI applications, today announced TTS-1.5. Inworld TTS-1.5 is a state-of-the-art text-to-speech (TTS) model that removes the critical barriers of latency, cost, and quality that have stalled innovation on realtime, consumer-scale AI applications.

For the past few years, the promise of a consumer AI revolution has remained largely unfulfilled. While enterprise AI has focused on cost-cutting, consumer applications have been hampered by slow, unengaging and costly user experiences. Inworld is addressing this by building a new, production-grade AI stack for consumer-scale applications from the ground up. Inworld TTS-1.5 is positioned to unblock voice-based AI experiences to reach millions of users, delivering the performance necessary for realtime at scale.

"The next wave of consumer-scale AI applications have been slow to scale because the current generation of models and infrastructure aren’t built for them. We've already been the number one model in benchmarks like Artificial Analysis, but with this release we’re addressing the remaining pain points we heard from developers" said Kylan Gibbs, CEO of Inworld.

"Inworld is building the missing model and infrastructure that will finally allow developers to create the kind of AI applications that people will actually want to use, and pay for. TTS-1.5 is the first step, proving that production-grade realtime latency, quality and cost is not only possible, but available today."

Inworld TTS-1.5: Unlocking Voice AI for the Next Wave of AI Applications

As the first release from Inworld's new production-grade AI stack for consumer-scale, TTS-1.5 delivers state-of-the-art performance, validated by industry benchmarks:

• Breakthrough Latency: With a P90 latency of 130ms (Mini model) and 250ms (Max model), TTS-1.5 is 4x faster than the previous generation, enabling truly real-time, interruptible conversations.
• Superior Quality and Expressiveness: A 40% improvement in word error rate and a 30% improvement in expressiveness result in more accurate, natural, and emotionally resonant speech.
• Radically Accessible Pricing: At just $5-10 per million characters, TTS-1.5 is more than 25x lower cost than the next best alternative.
  

Inworld's new infrastructure, beginning with TTS-1.5, is available starting today. Developers can learn more and request access at inworld.ai/tts.

“We chose Inworld because of its low latency, high-quality output, multilingual support and competitive pricing. In addition, with high rate limits and stable performance, Inworld has enabled us to scale confidently.” - Dimitri Dekanozishvili, Talkpal AI Co-Founder

“When we started working on Stellar Cafe two years ago, the #1 piece of feedback from playtests was that the voices weren't good enough. Some users were also skeptical that it would be possible for this to feel good. When we adopted Inworld TTS, it was a game changer. Immediately users switched and began mentioning how magical it was." – Devin Reimer, Astrobeam Founder/CEO.

"If you want top-of-the-line TTS integrated into your product, work with Inworld. It’s easy to use, quick, high-quality, and a fraction of the price of comparable services." - Sara Beykpour, Particle Co-founder & CEO

About Inworld AI

Inworld is a research lab developing AI models and infrastructure for the next wave of AI consumer-scale applications. With a focus on creating a production-grade AI stack, Inworld empowers developers to build more engaging and emotionally resonant experiences. The company was founded by a team of AI experts from Google and DeepMind, and is backed by leading investors like Lightspeed, Kleiner Perkins, and Stanford University.

Media Contact:

Andreas Assad
Head of Biz Ops
andreas@inworld.ai
617-543-4769

Legal Disclaimer:

EIN Presswire provides this news content "as is" without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author above.