DubGuild Secures Up to 2.1 Billion JPY in Funding and GENIAC Support to Accelerate 30B Spoken Language Model Development
Key facts
- DubGuild Secures Up to 2.1 Billion JPY in Funding and GENIAC Support to Accelerate 30B Spoken Language Model Development
- DubGuild Inc. has completed a seed round of third-party allotment funding led by JAFCO Group, with participation from Coreline Ventures and CyberAgent Capital. Concurrently, the company was selected for the 4th phase of "GENIAC," a generative AI development support project promoted by METI and NEDO. The combined funding and support amount to a maximum of approximately 2.1 billion JPY. The company will utilize the funds to accelerate the development and social implementation of a 30B-parameter end-to-end (E2E) Spoken Language Model (SLM) capable of natural conversation and emotional understanding without relying on text processing. Additionally, Professor Tomoki Toda of Nagoya University has joined as an advisor.
- Source: PR Times
- Date: June 6, 2026
Direct answer
DubGuild Inc. has completed a seed round of third-party allotment funding led by JAFCO Group, with participation from Coreline Ventures and CyberAgent Capital. Concurrently, the company was selected for the 4th phase of "GENIAC," a generative AI development support project promoted by METI and NEDO. The combined funding and support amount to a maximum of approximately 2.1 billion JPY. The company will utilize the funds to accelerate the development and social implementation of a 30B-parameter end-to-end (E2E) Spoken Language Model (SLM) capable of natural conversation and emotional understanding without relying on text processing. Additionally, Professor Tomoki Toda of Nagoya University has joined as an advisor.
- Citation
- DubGuild Secures Up to 2.1 Billion JPY in Funding and GENIAC Support to Accelerate 30B Spoken Language Model Development (June 6, 2026), PR Times
- Source
- PR Times
- Date
- June 6, 2026
DubGuild Inc. has completed a seed round of third-party allotment funding led by JAFCO Group, with participation from Coreline Ventures and CyberAgent Capital. Concurrently, the company was selected for the 4th phase of "GENIAC," a generative AI development support project promoted by METI and NEDO. The combined funding and support amount to a maximum of approximately 2.1 billion JPY. The company will utilize the funds to accelerate the development and social implementation of a 30B-parameter end-to-end (E2E) Spoken Language Model (SLM) capable of natural conversation and emotional understanding without relying on text processing. Additionally, Professor Tomoki Toda of Nagoya University has joined as an advisor.
📋 Article Processing Timeline
- 📰 Published: June 6, 2026 at 00:00
- 🔍 Collected: June 5, 2026 at 15:20
- 🤖 AI Analyzed: June 6, 2026 at 13:18 (21h 57m after Collected)
DubGuild Inc. (Headquarters: Bunkyo-ku, Tokyo; CEO: Masatoshi Ohtake; hereinafter "DubGuild") announced the completion of a seed round of funding through a third-party allotment. The round was led by JAFCO Group Co., Ltd. (hereinafter "JAFCO"), with participation from Coreline Ventures and CyberAgent Capital.
Simultaneously, DubGuild announced that its proposal, "Development and Social Implementation of a Spoken Foundation Model for Human-like Dialogue," has been selected for the 4th phase of "GENIAC" (Generative AI Accelerator Challenge), a project promoted by the Ministry of Economy, Trade and Industry (METI) and the New Energy and Industrial Technology Development Organization (NEDO) to strengthen domestic generative AI development capabilities. With this funding and various support programs, the total accumulated amount of funding and support reaches a maximum of approximately 2.1 billion JPY.
With this procurement, DubGuild will strengthen its recruitment, focusing primarily on engineers, and further accelerate the advancement and social implementation of the non-text-mediated, voice-specialized foundation model currently under research and development.
Company Overview
DubGuild operates under the mission of "Scaling Speech AI" and engages in research and development to create the world's most outstanding spoken language foundation models. Previous R&D initiatives are published on the company's blog.
Company Name: DubGuild Inc.
Representative: Masatoshi Ohtake, CEO
Location: East U. HiRAKUGATE, 2-3-10 Mukougaoka, Bunkyo-ku, Tokyo 113-0023
Website: https://dubguild.com
Business: R&D of spoken dialogue AI foundation models and generative AI technologies
Dev Blog: https://blog.dubguild.com/melte/llm-tts-scaling
Development Overview
The selected proposal involves developing a 30B-parameter spoken language foundation model that enables human-like interaction. The goal is to realize real-time dialogue encompassing emotional understanding, natural responses, and the ability to handle interruptions.
NEDO Publication: https://www.nedo.go.jp/content/800056193.pdf
Background and Purpose
In recent years, Large Language Models (LLMs) have garnered significant attention for their text-based conversational abilities, vastly improving productivity. However, since humans primarily communicate via voice, there is a strong demand to transition to voice-based models. The most naive approach to achieving this is utilizing a pipeline of "Automatic Speech Recognition (ASR) + LLM + Text-to-Speech (TTS)." Yet, this method has limitations regarding emotional expression and real-time responsiveness, creating barriers to natural conversation and widespread adoption. Under these circumstances, end-to-end (E2E) spoken dialogue models are emerging, and related research is accelerating globally. Against this backdrop, DubGuild aims to develop a 30B-parameter Spoken Language Model (SLM) with an eye on acquiring international competitiveness and leading the R&D of E2E spoken dialogue AI.
About GENIAC
GENIAC is a project promoted by METI and NEDO designed to bolster domestic generative AI development capabilities. It provides support such as securing computational resources, accumulating datasets, and sharing knowledge necessary for foundation model development.
URL: https://www.meti.go.jp/policy/mono_info_service/geniac/index.html
Appointment of Professor Tomoki Toda as Advisor
To further propel the DubGuild model R&D business, Professor Tomoki Toda has been appointed as an advisor to the company as of May 2026.
Professor Toda is currently a professor at the Institutes of Innovation for Future Society and the Graduate School of Informatics at Nagoya University, specializing in speech information processing. He is a leading researcher in spoken media generation and conversion technologies, including speech synthesis, speech analysis, and music information processing.
Profile URL: https://sites.google.com/site/tomokitoda
Representative Comment
Masatoshi Ohtake, CEO of DubGuild Inc.
"As foundation models profoundly alter the course of humanity, I want us to be on the side that creates these changes, not just the side that receives them. Japan possesses the capability to build foundation models, and I want to prove that through this company. The voice domain we are currently focusing on is central to human activity and an area where Japan has strong advantages. Let's do this together!"
Investor Comments
Yutaro Saka / Takumi Nakamura, Partners at JAFCO
"We have been watching voice AI as an area with expanding possibilities. DubGuild is directly tackling the formidable challenge of developing models from scratch. It is not an easy path, but they have assembled an incredibly attractive team with the technical prowess and passion required to succeed. We believe they will create a globally competitive product originating from Japan, and we are thrilled to accompany them on this journey."
Kenichiro Hara / Yojiro Yamamoto, GP & Co-founders at Coreline Ventures
"Following our lead investment in 2025, we have made an additional investment in this round. While the US and China dominate LLM development, Japan has consistently maintained world-leading laboratories in speech processing and synthesis. We believe that in the voice-specialized domain of foundation models, Japanese companies have a chance to compete globally. DubGuild is composed of alumni from these top labs, led by CEO Ohtake. We welcome anyone who wants to join or support this passionate and talented group!"
Keita Yazaki, Senior Associate at CyberAgent Capital, Inc.
"We were given the opportunity to invest in DubGuild this time. Speech synthesis is a field where Japan has led the world, and there are still significant opportunities left in generating rich voice expressions. I felt that the development team led by Mr. Ohtake could realize a voice intelligence model representing Japan's speech synthesis and enabling the richest expressions. We encourage researchers and anyone interested in this challenge to contact us."
About Recruitment
DubGuild is currently actively recruiting engineers responsible for voice AI, machine learning, and product development.
Simultaneously, DubGuild announced that its proposal, "Development and Social Implementation of a Spoken Foundation Model for Human-like Dialogue," has been selected for the 4th phase of "GENIAC" (Generative AI Accelerator Challenge), a project promoted by the Ministry of Economy, Trade and Industry (METI) and the New Energy and Industrial Technology Development Organization (NEDO) to strengthen domestic generative AI development capabilities. With this funding and various support programs, the total accumulated amount of funding and support reaches a maximum of approximately 2.1 billion JPY.
With this procurement, DubGuild will strengthen its recruitment, focusing primarily on engineers, and further accelerate the advancement and social implementation of the non-text-mediated, voice-specialized foundation model currently under research and development.
Company Overview
DubGuild operates under the mission of "Scaling Speech AI" and engages in research and development to create the world's most outstanding spoken language foundation models. Previous R&D initiatives are published on the company's blog.
Company Name: DubGuild Inc.
Representative: Masatoshi Ohtake, CEO
Location: East U. HiRAKUGATE, 2-3-10 Mukougaoka, Bunkyo-ku, Tokyo 113-0023
Website: https://dubguild.com
Business: R&D of spoken dialogue AI foundation models and generative AI technologies
Dev Blog: https://blog.dubguild.com/melte/llm-tts-scaling
Development Overview
The selected proposal involves developing a 30B-parameter spoken language foundation model that enables human-like interaction. The goal is to realize real-time dialogue encompassing emotional understanding, natural responses, and the ability to handle interruptions.
NEDO Publication: https://www.nedo.go.jp/content/800056193.pdf
Background and Purpose
In recent years, Large Language Models (LLMs) have garnered significant attention for their text-based conversational abilities, vastly improving productivity. However, since humans primarily communicate via voice, there is a strong demand to transition to voice-based models. The most naive approach to achieving this is utilizing a pipeline of "Automatic Speech Recognition (ASR) + LLM + Text-to-Speech (TTS)." Yet, this method has limitations regarding emotional expression and real-time responsiveness, creating barriers to natural conversation and widespread adoption. Under these circumstances, end-to-end (E2E) spoken dialogue models are emerging, and related research is accelerating globally. Against this backdrop, DubGuild aims to develop a 30B-parameter Spoken Language Model (SLM) with an eye on acquiring international competitiveness and leading the R&D of E2E spoken dialogue AI.
About GENIAC
GENIAC is a project promoted by METI and NEDO designed to bolster domestic generative AI development capabilities. It provides support such as securing computational resources, accumulating datasets, and sharing knowledge necessary for foundation model development.
URL: https://www.meti.go.jp/policy/mono_info_service/geniac/index.html
Appointment of Professor Tomoki Toda as Advisor
To further propel the DubGuild model R&D business, Professor Tomoki Toda has been appointed as an advisor to the company as of May 2026.
Professor Toda is currently a professor at the Institutes of Innovation for Future Society and the Graduate School of Informatics at Nagoya University, specializing in speech information processing. He is a leading researcher in spoken media generation and conversion technologies, including speech synthesis, speech analysis, and music information processing.
Profile URL: https://sites.google.com/site/tomokitoda
Representative Comment
Masatoshi Ohtake, CEO of DubGuild Inc.
"As foundation models profoundly alter the course of humanity, I want us to be on the side that creates these changes, not just the side that receives them. Japan possesses the capability to build foundation models, and I want to prove that through this company. The voice domain we are currently focusing on is central to human activity and an area where Japan has strong advantages. Let's do this together!"
Investor Comments
Yutaro Saka / Takumi Nakamura, Partners at JAFCO
"We have been watching voice AI as an area with expanding possibilities. DubGuild is directly tackling the formidable challenge of developing models from scratch. It is not an easy path, but they have assembled an incredibly attractive team with the technical prowess and passion required to succeed. We believe they will create a globally competitive product originating from Japan, and we are thrilled to accompany them on this journey."
Kenichiro Hara / Yojiro Yamamoto, GP & Co-founders at Coreline Ventures
"Following our lead investment in 2025, we have made an additional investment in this round. While the US and China dominate LLM development, Japan has consistently maintained world-leading laboratories in speech processing and synthesis. We believe that in the voice-specialized domain of foundation models, Japanese companies have a chance to compete globally. DubGuild is composed of alumni from these top labs, led by CEO Ohtake. We welcome anyone who wants to join or support this passionate and talented group!"
Keita Yazaki, Senior Associate at CyberAgent Capital, Inc.
"We were given the opportunity to invest in DubGuild this time. Speech synthesis is a field where Japan has led the world, and there are still significant opportunities left in generating rich voice expressions. I felt that the development team led by Mr. Ohtake could realize a voice intelligence model representing Japan's speech synthesis and enabling the richest expressions. We encourage researchers and anyone interested in this challenge to contact us."
About Recruitment
DubGuild is currently actively recruiting engineers responsible for voice AI, machine learning, and product development.
FAQ
What is the total funding and support amount for DubGuild?
The combined total from their seed funding round and support programs like GENIAC reaches up to approximately 2.1 billion JPY.
What makes DubGuild's spoken language model unique?
Unlike traditional pipeline methods, it uses an end-to-end (E2E) architecture, enabling emotion recognition and natural real-time conversations including interruptions without converting to text.
What is the GENIAC project?
GENIAC is an initiative promoted by the Japanese Ministry of Economy, Trade and Industry (METI) and NEDO to strengthen domestic generative AI development capabilities.