Palabra.ai takes #1 spot for speed in new text-to-speech (TTS) benchmark. PALABRA.AI LTD Jul 31, 2026, 10:41 ET Palabra ranks #1 for latency on Coval's independent, open-source text-to-speech benchmark, posting 104 milliseconds - roughly twice as fast as the nearest competitor - with a 6% word error rate. Coval's independent benchmark puts Palabra's text-to-speech model at 104 ms latency - about twice as fast as the next-closest competitors. LONDON, July 31, 2026 /PRNewswire-PRWeb/ - Palabra, a real-time speech AI company, has ranked #1 for latency on Coval's independent TTS benchmark, a widely watched, continuously updated leaderboard tracking how leading voice AI models perform under real-world conditions. Palabra posted 104 milliseconds of latency and a 6% word error rate (WER), outpacing ElevenLabs, Cartesia, and other major players in the space - roughly twice as fast as the nearest competitor. Latency is one of the key factors determining whether a voice AI system can support a fluid, real-time conversation. Even when the generated voice itself sounds highly natural, a noticeable delay before each response can disrupt the flow of dialogue, creating awkward pauses, causing people to talk over each other, and making the interaction feel less immediate. Many voice AI systems still operate at around 500 milliseconds of latency or more. Palabra's model achieves 104 milliseconds in Coval's independent benchmark, significantly reducing the delay between turns and enabling a more fluid, uninterrupted dialogue. Coval's benchmark is built to reflect production reality rather than idealized lab conditions: it measures Time to First Audio (TTFA) - the delay a listener actually notices, including any silence before the first audible sample - using pinned, versioned datasets so every provider is scored against identical inputs. Word error rate is calculated by transcribing each provider's synthesized audio with a fixed ASR model and scoring it against the original text, so the number reflects whether the speech is intelligible, not just fast. The full runner and methodology are open-source and independently reproducible. "Latency is one of the most consequential metrics in voice AI," said Brooke Hopkins, Founder and CEO of Coval. "Humans respond in around 400 milliseconds, and systems slower than that feel unnatural regardless of output quality. What makes Palabra's result significant is what it creates downstream: when a model is this fast, every other component in the stack gets more time to think." Palabra says the result reflects a new TTS architecture that achieves 35ms of time-to-first-audio before network overhead, paired with production infrastructure built to hold that speed at scale across large volumes of concurrent users. "The biggest unsolved challenge that the world's leading speech AI labs are working on today is making interactions with AI voice agents feel indistinguishable from conversations with another person," said Artem Kukharenko, CEO & Co-founder of Palabra Al. "Latency above 200 milliseconds introduces a noticeable delay that disrupts the natural flow of communication. That's why we've focused on building TTS with extremely low time-to-first-audio that could work at large scale in production systems. In Coval's independent benchmark, we rank #1 at 104ms - the closest competitor is approximately twice as slow." The TTS result is part of a broader focus at Palabra on real-time speech models - spanning text-to-speech, speech recognition (ASR), and speech-to-speech translation - built around low latency as a core design principle across the stack. About Palabra Palabra is a voice AI lab developing real-time models for text-to-speech (TTS), automatic speech recognition (ASR), and speech-to-speech translation. Its proprietary models and APIs enable developers and enterprises to build low-latency multilingual voice experiences across customer support, conferencing, live streaming, education, and other real-time applications. Palabra recently raised $8.4 million in pre-seed funding led by Seven Seven Six (776) with participation from Creator Ventures, and prominent angel investors including Max Mullen, co-founder of Instacart; Anne Lee Skates, former partner at Andreessen Horowitz; Mehdi Ghissassi,, former Head of Product at DeepMind; and Namat Bahram, an early backer of ElevenLabs. About Coval Coval is the San Francisco-based evaluation platform for voice AI, founded by ex-Waymo engineer Brooke Hopkins. It builds independent, continuously updated, open-source benchmarks that measure how voice AI systems perform in production, rather than in idealized lab conditions. Coval recently raised a $28M Series A led by Norwest, with Twilio Ventures and Y Combinator also on the cap table, and is already trusted by over 60 organizations, including Zoom and Deepgram. Its public TTS leaderboard is available at benchmarks.coval.ai/tts. Media Contact Dmytro Tymoshenko, PALABRA.AI LTD, 380 939944325, [email protected], palabra.ai SOURCE PALABRA.AI LTD

Palabra.ai takes #1 spot for speed in new text-to-speech (TTS) benchmark. Business · JUL 31, 2026 PR Newswire Palabra ranks #1 for latency on Coval's independent, open-source text-to-speech benchmark, posting 104 milliseconds - roughly twice as fast as the nearest... Palabra ranks #1 for latency on Coval's independent, open-source text-to-speech benchmark, posting 104 milliseconds - roughly twice as fast as the nearest competitor - with a 6% word error rate. Coval's independent benchmark puts Palabra's text-to-speech model at 104 ms latency - about twice as fast as the next-closest competitors. LONDON, July 31, 2026 /PRNewswire-PRWeb/ - Palabra, a real-time speech AI company, has ranked #1 for latency on Coval's independent TTS benchmark, a widely watched, continuously updated leaderboard tracking how leading voice AI models perform under real-world conditions. Palabra posted 104 milliseconds of latency and a 6% word error rate (WER), outpacing ElevenLabs, Cartesia, and other major players in the space - roughly twice as fast as the nearest competitor. The hardest problem in speech AI is making voice agents feel indistinguishable from a real person. Latency above 200 milliseconds breaks that. We built our TTS for extremely low time-to-first-audio at production scale. Coval's benchmark ranks us #1 at 104 ms, nearly twice as fast as the next model. Latency is one of the key factors determining whether a voice AI system can support a fluid, real-time conversation. Even when the generated voice itself sounds highly natural, a noticeable delay before each response can disrupt the flow of dialogue, creating awkward pauses, causing people to talk over each other, and making the interaction feel less immediate. Many voice AI systems still operate at around 500 milliseconds of latency or more. Palabra's model achieves 104 milliseconds in Coval's independent benchmark, significantly reducing the delay between turns and enabling a more fluid, uninterrupted dialogue. Coval's benchmark is built to reflect production reality rather than idealized lab conditions: it measures Time to First Audio (TTFA) - the delay a listener actually notices, including any silence before the first audible sample - using pinned, versioned datasets so every provider is scored against identical inputs. Word error rate is calculated by transcribing each provider's synthesized audio with a fixed ASR model and scoring it against the original text, so the number reflects whether the speech is intelligible, not just fast. The full runner and methodology are open-source and independently reproducible. "Latency is one of the most consequential metrics in voice AI," said Brooke Hopkins, Founder and CEO of Coval. "Humans respond in around 400 milliseconds, and systems slower than that feel unnatural regardless of output quality. What makes Palabra's result significant is what it creates downstream: when a model is this fast, every other component in the stack gets more time to think." Palabra says the result reflects a new TTS architecture that achieves 35ms of time-to-first-audio before network overhead, paired with production infrastructure built to hold that speed at scale across large volumes of concurrent users. "The biggest unsolved challenge that the world's leading speech AI labs are working on today is making interactions with AI voice agents feel indistinguishable from conversations with another person," said Artem Kukharenko, CEO & Co-founder of Palabra Al. "Latency above 200 milliseconds introduces a noticeable delay that disrupts the natural flow of communication. That's why we've focused on building TTS with extremely low time-to-first-audio that could work at large scale in production systems. In Coval's independent benchmark, we rank #1 at 104ms - the closest competitor is approximately twice as slow." The TTS result is part of a broader focus at Palabra on real-time speech models - spanning text-to-speech, speech recognition (ASR), and speech-to-speech translation - built around low latency as a core design principle across the stack. About Palabra Palabra is a voice AI lab developing real-time models for text-to-speech (TTS), automatic speech recognition (ASR), and speech-to-speech translation. Its proprietary models and APIs enable developers and enterprises to build low-latency multilingual voice experiences across customer support, conferencing, live streaming, education, and other real-time applications. Palabra recently raised $8.4 million in pre-seed funding led by Seven Seven Six (776) with participation from Creator Ventures, and prominent angel investors including Max Mullen, co-founder of Instacart; Anne Lee Skates, former partner at Andreessen Horowitz; Mehdi Ghissassi,, former Head of Product at DeepMind; and Namat Bahram, an early backer of ElevenLabs. About Coval Coval is the San Francisco-based evaluation platform for voice AI, founded by ex-Waymo engineer Brooke Hopkins. It builds independent, continuously updated, open-source benchmarks that measure how voice AI systems perform in production, rather than in idealized lab conditions. Coval recently raised a $28M Series A led by Norwest, with Twilio Ventures and Y Combinator also on the cap table, and is already trusted by over 60 organizations, including Zoom and Deepgram. Its public TTS leaderboard is available at benchmarks.coval.ai/tts. Media Contact Dmytro Tymoshenko, PALABRA.AI LTD, 380 939944325, [email protected], palabra.ai SOURCE PALABRA.AI LTD Published by News Desk · WeeklyReviewer The WeeklyReviewer news desk monitors breaking developments around the clock - sourcing, verifying, and publishing real-time industry news across business, technology, politics, science, sports, and world affairs. Every story that comes through the Live Wire is reviewed for accuracy before publication, keeping our readers ahead of the curve without the noise.

Tysa turns one: Conectys' AI voice agent marks a Year of Global conversation. Conectys: AI-Powered Business Transformation Outsourcing Tysa: one year of meaningful interactions, powered by intelligence. Tysa's first year signals a shift in global CX, enabling scalable, real-time multilingual voice AI interactions with AI and human-in-the-loop design. BUCHAREST, ROMANIA, July 30, 2026 / EINPresswire.com / - Conectys, the global Business Transformation Outsourcing (BTO) leader, marks the first anniversary of Tysa, its next-generation conversational AI voice agent. Since its launch in June 2025, the solution has powered thousands of conversations across 55 countries and in more than 20 languages. This reflects Conectys' shift from traditional outsourcing to an AI-native, people-powered BTO model built for global operations and multilingual audiences. A Year of Conversations That Matter One year ago, Conectys set out with a clear goal: to make it easier for people to connect with their preferred brands and find the information they need, naturally and in their own language. Today, that ambition has materialized into a globally deployed voice AI agent handling inbound and outbound interactions around the clock, across time zones and cultures. This direction aligns with a broader market shift, as the voice AI agent market is projected to grow at a compound annual rate of 34.8%, from $2.4 billion in 2024 to $47.5 billion by 2034 (Market.us). Companies are turning to solutions like Tysa for more than operational cost savings. AI voice agents can take care of a large share of routine customer conversations, helping teams respond faster and provide support around the clock. This allows human agents to focus on the more complex situations where empathy, judgment, and problem-solving matter most. The result is often a better customer experience, greater operational efficiency, and a support model that can scale as customer expectations continue to grow. "Tysa is not just a product innovation. It is a demonstration of who Conectys is becoming," said Tomasz Kacała, Data Solutions Manager at Conectys. "The intersection of human experience and intelligent automation is where we build lasting value for our clients. Tysa embodies exactly that." What Tysa Is and What Sets It Apart Built by the IT Data Solutions team at Conectys in partnership with ElevenLabs, Tysa goes beyond traditional IVR systems. Instead of rigid menus and scripted replies, it enables fluid, human-like conversations powered by speech-to-text, LLM reasoning, and advanced text-to-speech. The result is a natural, expressive voice experience that adapts in real time to context, similar to a skilled human agent. Tysa's key capabilities include: 1. Inbound Voice Conversations: Tysa answers calls, guides users step by step, and captures key information seamlessly, handling real conversations and tasks at scale. 2. Intelligent Workflow Guidance: Rather than following a fixed script, Tysa adapts each response based on the context of the interaction, functioning as a true voice assistant that walks users through complex processes. 3. Agentic AI Workflows: Tysa securely connects to external systems to act mid-conversation, scheduling meetings, updating records, and triggering downstream workflows autonomously. 4. Outbound Voice Calls: Tysa proactively contacts customers to conduct follow-ups, surveys, and confirmations, all in a natural, human tone. 5. Human-in-the-Loop by Design: When interactions require empathy, judgment, or nuance beyond its scope, Tysa seamlessly transfers the conversation to a human agent, with full context transferred so the customer never has to repeat themselves. 6. Enterprise-Grade Responsible AI: Every interaction runs on a secure, compliant, fully governed AI infrastructure, ensuring Tysa can be trusted to represent a brand at scale. Real-World Use Cases: How Tysa Scales Voice CX Across Key Industries For global brands, the value of voice AI lies in its ability to handle high-volume, time-sensitive interactions with speed, consistency, and cultural reach. Tysa is especially well suited to industries where expectations are high, demand fluctuates quickly, and seamless escalation to human support remains essential. Here are a few examples: In gaming, Tysa can support players with account access, in-game purchases, and safety-related reports, providing studios always-on coverage while human teams focus on sensitive moderation work. In e-commerce, Tysa can manage high volumes of inquiries about orders, deliveries, returns, and product information, resolving routine requests autonomously and passing nuanced cases to CX agents with full context. In hospitality, Tysa can assist customers with bookings, itinerary changes, and real-time updates on delays or cancellations, helping airlines and hospitality brands provide consistent, multilingual voice support across time zones. Why Tysa Reflects Conectys' BTO Vision Tysa is not an isolated product. It reflects Conectys' shift from Business Process Outsourcing (BPO) to Business Transformation Outsourcing (BTO), where AI operates alongside human talent as part of the delivery model. Conectys' integrated talent engine brings together four complementary strengths: 1. Multilingual agents handling high-empathy interactions. 2. Native AI solutions operating autonomously, where Tysa adds real value. 3. A global gig workforce that scales on demand. 4. Employer of Record (EoR) expertise enabling compliant global hiring. The Road Ahead One year in, Tysa continues to evolve. Conectys will expand its language capabilities, deepen workflow integrations, and explore new use cases across client verticals, including contact center support, data annotation QA, Trust & Safety, and outbound engagement programs. "AI powers efficiency. Humans bring empathy and insight. Together, they create experiences people trust," added Tomasz Kacała. About Conectys Conectys is a global Business Transformation Outsourcing (BTO) company specializing in multilingual customer experience, content moderation, and data annotation services. With more than 20 years of experience, it has evolved from a regional outsourcing provider into a BTO partner for multinational brands, working across sectors such as retail, e-commerce, travel, fintech, gaming, social media, and more. Today, Conectys delivers solutions in more than 100 languages globally, operating 14 modern delivery hubs and a global gig workforce in over 180 countries. By combining skilled people, AI expertise, and elastic global capacity, Conectys helps clients launch, adapt, and scale complex operations with speed and confidence. About Tomasz Kacała Tomasz Kacała is Data Solutions Manager at Conectys, where he works alongside the IT Data Solutions team on the data and AI initiatives powering the company's Business Transformation Outsourcing (BTO) model, including the development of Tysa. Fawzia Safdari, Vice President of Global Marketing Conectys +1 954-600-1267 email us here Visit us on social media: LinkedIn YouTube Tysa by Conectys: Redefining Conversational CX 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.