MOUNTAIN VIEW, Calif., Jan. 21, 2026 (Globe Newswire) — Jan. 21, 2026 – Inworld AI, a research organization 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 significant barriers of latency, cost, and quality that have slowed innovation in real-time, consumer-scale AI applications.
Over the past few years, the promise of the consumer AI revolution has gone largely unfulfilled. While enterprise AI is focused on reducing costs, consumer applications have been hampered by slow, unappealing, and costly user experiences. Inworld is addressing this issue by building a new production-grade AI stack from the ground up for consumer-scale applications. Inworld TTS-1.5 unblocks voice-based AI experiences to reach millions of users and provides the performance needed for real-time at scale.
“The next wave of consumer-scale AI applications has been slow to scale because current-generation models and infrastructure weren’t built for them. We’re already the number-one model in benchmarks like Artificial Analysis, and this release addresses the remaining pain points we’ve heard from developers,” said Kylan Gibbs, CEO of Inworld.
“Inworld is building the missing models and infrastructure that will enable developers to create the kinds of AI applications that people actually want to use and pay for. TTS-1.5 is the first step in that, proving that production-grade real-time latency, quality, and cost is not only possible, but available today.”
Inworld TTS-1.5: Unleashing voice AI for the next wave of AI applications
TTS-1.5, the first release of Inworld’s new consumer-scale production-grade AI stack, delivers cutting-edge performance validated by industry benchmarks.
- breakthrough latency: With P90 latency of 130ms (mini model), 250ms (maximum model), TTS-1.5 is 4x faster than the previous generation and enables truly real-time, interruptible conversations.
- Excellent quality and expressiveness: Improved word error rate by 40% and expressiveness by 30%, resulting in more accurate, natural, and emotionally resonant speech.
- Thoroughly accessible pricing: At only $5 to $10 per million characters, TTS-1.5 is more than 25 times cheaper than the next best alternative.
Inworld’s new infrastructure, including TTS-1.5, is available starting today. Developers can learn more and request access. inworld.ai/tts.
“We chose Inworld for its low latency, high-quality output, multilingual support, and competitive pricing. Additionally, Inworld allows us to scale with confidence with high rate limits and consistent performance.” – Dimitri Dekanozhishvili, Co-Founder of Talkpal AI
“When we started working on Stella Cafe two years ago, the most common feedback we received during playtesting was that the audio wasn’t good enough. Some users were also skeptical that they would feel comfortable with it. When we adopted Inworld TTS, it was a game changer. Immediately users started switching and saying how magical it was.” – Devin Reimer, Astrobeam Founder/CEO.
“If you want to integrate top-of-the-line TTS into your product, work with Inworld. Easy to use, fast, high quality, and at a fraction of the price of comparable services.” – Sara Beykpour, Particle Co-Founder and CEO
About inworld AI
Inworld is a research institute developing AI models and infrastructure for the next wave of consumer-scale AI applications. Inworld is focused on creating production-grade AI stacks that enable 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 major investors including Lightspeed, Kleiner Perkins, and Stanford University.
Media contact:
andreas assad
Head of Business Operations
andreas@inworld.ai
617-543-4769
