Kotoba Raises $10M to Power the Next Wave of Voice AI in East Asia

SAN FRANCISCO & TOKYO – June 24, 2026 – Kotoba Technologies, a developer of hyper-specialized voice AI, has secured an additional $10 million in seed funding to expand its real-time speech platform across East Asia. The round, which brings the company's total funding to $23 million, was led by Kindred Ventures and saw participation from strategic heavyweights Salesforce Ventures and Sony Innovation Fund.

The investment signals strong confidence in Kotoba's mission to solve one of AI’s most complex challenges: understanding and generating nuanced, real-time speech in East Asian languages. The capital will fuel the advancement of its proprietary Koto model, focusing on speech-to-speech translation, on-device processing for smart hardware, and a broader rollout of its agentic AI solutions for enterprise clients.

Founded in 2023 by AI researchers Dr. Noriyuki Kojima (CEO) and Dr. Jungo Kasai (CTO), the company has rapidly distinguished itself by creating models that it claims outperform those from larger American and European labs in both speech recognition and synthesis for Japanese, Korean, and Chinese.

Cracking the Code of Complex Languages

While many large language models aim for global coverage, they often fall short in capturing the intricate grammatical structures, tonal variations, and cultural contexts of East Asian languages. Kotoba Technologies has staked its claim by focusing intensely on this underserved but massive market. Its core technology, the Koto model, is an end-to-end system purpose-built for the demands of real-time applications, from AI agents to simultaneous translation.

"The complexities of getting the unique aspects of Asian languages requires a unique training strategy and learning loop approach with a deep understanding of each language and market," said Steve Jang, Founder and Managing Partner at Kindred Ventures. He noted that Kotoba's models "performed better than existing models developed by American and European research labs" on both recognition and synthesis.

The company’s platform offers a suite of flexible models, including speech-to-speech (S2S) systems that have demonstrated simultaneous translation with sub-two-second latency. It also provides ultra-low-latency speech-to-text (ASR) and text-to-speech (TTS) models, which can be deployed either in the cloud or directly on edge devices like smartphones and wearables. This on-device capability is critical for privacy, speed, and enabling a new generation of smart, responsive hardware.

This technology is not just a lab experiment. The company's consumer-facing app, “Kotoba,” has already surpassed 180,000 users. The app provides real-time, high-quality translation across 21 languages and includes productivity features like AI-powered note-taking and summaries. A major update in June introduced 11 new features, and a meeting-agent experience for remote conferences is slated for July, further bridging the gap between consumer and enterprise applications.

Strategic Capital for a Foundational Platform

The investor lineup reveals a strategic vision that extends far beyond a simple financial return. The participation of Salesforce Ventures and Sony Innovation Fund points to Kotoba’s potential as a foundational infrastructure layer for both enterprise software and consumer electronics.

Salesforce, a global leader in CRM, sees a clear path to integrating Kotoba's capabilities into its ecosystem. "In addition to their high technical capabilities, we see immense potential in their focus on driving implementation in business environments," noted Ken Asada, a Partner at Salesforce Ventures. The investment aligns with Salesforce's strategy of embedding sophisticated AI to help its global customers operate more effectively, especially in the booming East Asian market.

Meanwhile, Sony's involvement underscores the technology's relevance for the future of hardware. Kotoba’s plan to dedicate resources to running its Koto model efficiently on edge chips for automobiles, electronics, and AI wearables directly complements Sony’s interests. "Real-time voice communication remains one of the most technically challenging AI frontiers," said Austin Noronha, Managing Director at Sony Ventures-US. "Kotoba is building more than a translation application, it is creating a voice AI infrastructure platform with potential applications across enterprise, telecom, electronics, and consumer markets."

Beyond Translation: Building the Future of Conversational AI

Kotoba is positioning its technology not merely as a translation tool but as the engine for the next generation of conversational AI. A key part of its strategy is the "agentic rollout," aimed at improving the usability of the Koto ecosystem for enterprises looking to deploy sophisticated AI voice agents and expand into Asian markets.

To accelerate this vision, the company has released an alpha version of its API and a Python SDK, opening its platform to a wider developer community. This allows external teams to build on Kotoba's S2S, ASR, and TTS models, testing them via the API before committing to on-device deployments. This platform-based approach is already bearing fruit, with the Koto model in production with Fortune Global 500 companies and AI-native startups. Use cases range from powering voice interfaces in contact centers to enabling AI-powered simultaneous translation for wearable devices. One notable partnership includes SoftBank, which features the Koto model in its "Daredemo AI" service, demonstrating significant enterprise validation.

Navigating a Complex Regulatory Landscape

Operating at the cutting edge of AI in East Asia requires navigating a complex web of regulatory and ethical considerations. Data privacy laws, such as Japan’s APPI, South Korea’s PIPA, and China’s PIPL, impose strict rules on handling personal data, including voice recordings. Kotoba appears to be addressing this proactively. Its ISO/IEC 27001:2022 certification for its Japan operations demonstrates a commitment to information security.

Furthermore, the company's privacy policy distinguishes between individual user data, which may be used for model improvement with an opt-out, and enterprise customer data, which is used solely for service provision. This separation is crucial for winning the trust of large corporations concerned about data security and intellectual property. The emphasis on on-device processing also provides a powerful privacy advantage, as it minimizes the need to send sensitive voice data to the cloud, helping to comply with stringent data localization laws and user expectations for privacy in the AI era.