Updated:
MetaReal
MetaReal holds Japan's #1 AI translation market share with its T-4OO engine across 2,000 domains and 6,000+ corporate users.
MetaReal
MetaReal operates as a Tokyo-based applied-AI product house that bundles generative AI, automated translation, digital twins and VR under a single corporate umbrella. Its public mission — liberating people from constraints of place, time and language — is executed through a narrow commercial thesis: producing industry-specific AI modules that plug directly into enterprise workflows. The firm is listed on the Tokyo Stock Exchange and its financial disclosures note that subsidiary Rosetta's translation engine generates the English-language versions of its earnings releases, an operational signal of eating its own cooking. The company organizes its offering around four pillars: the proprietary T-4OO neural translation engine covering 100 languages and 2,000 specialized domains; the "Shigoto Owaru AI" series of 50-plus industry-specific generative-AI tools; Metareal AI, a fully custom development service for enterprise clients; and a metaverse/digital-twin practice under the STUDIO55 and Gaussian Splatting labels. Deployment spans legal, pharmaceutical, financial, chemical, IT, machinery and electronics sectors. Co-development partnerships with Toyo Keizai (the "Shikiho API" for multi-angle business analysis in finance) and Digital Hearts (the "ella" character-translation engine for gaming) demonstrate a repeatable model of embedding AI into existing industry data products. The firm serves more than 6,000 clients, and ITR's "ITR Market View: Conversational AI & Machine Learning Platform Market 2024" ranks it first by vendor revenue share in the Japanese translation market (per ITR, 2024). Its generative-AI tools target functions such as PR writing and scoring (Koho AI) and pharmaceutical regulatory-document review (RakuYaku AI). Real-time speech translation for web conferencing is delivered through the "OnYaku" device-agnostic service. MetaReal's architecture hinges on the tight coupling of its proprietary language model with a vertical-industry distribution strategy — it does not offer a generic large language model but instead sells domain-tuned modules that sit on top of existing enterprise systems. This makes it a candidate for entities seeking to allocate to Asia-based AI infrastructure without taking pure model risk. Governance and leadership details are not publicly profiled on the firm's website, creating an information gap for external allocators.
General information
Firm type
Asset Manager
Year founded
—
AUM
Undisclosed
Location
Region
Asia
Country
Japan
City
—
Corporate office
Japan
Sector focus
Frequently asked questions
What does MetaReal actually sell — is it a foundation model company or an application-layer AI developer?
MetaReal is an application-layer developer that sells industry-specific AI modules, not a general-purpose foundation model. Its core asset is the proprietary neural translation engine T-4OO, which it wraps into pre-built tools — such as "Koho AI" for PR professionals and "RakuYaku AI" for pharmaceutical document review — or deploys via custom development. It does not market a standalone large language model to compete with OpenAI or Anthropic.
Which industries drive MetaReal's revenue, and how is that diversification structured?
The firm discloses active deployments across legal, pharmaceutical, financial, chemical, IT, machinery and electronics sectors. Its industry coverage is delivered through a library of more than 50 specialized products under the "Shigoto Owaru AI" series, each tuned for a specific vertical workflow. Co-developed products like the Toyo Keizai "Shikiho API" for financial analysis and the Digital Hearts "ella" engine for game-character translation provide additional revenue channels.
How does MetaReal's translation engine differ from off-the-shelf machine translation tools used by global enterprises?
The T-4OO engine is trained on 2,000 specialized domains — including legal, medical and engineering terminology — and MetaReal claims 95% accuracy using a neural-network architecture. The firm emphasizes file-level processing that handles scanned PDFs without pre-formatting, a feature designed to integrate into document-heavy Japanese enterprise workflows. ITR's 2024 market report ranked MetaReal first in Japanese translation vendor revenue share (per ITR, 2024).
Is MetaReal a public company, and what does that imply for an institutional allocation?
MetaReal is listed on the Tokyo Stock Exchange. For institutional allocators, this means exposure would be through public equities rather than a private fund commitment. The firm's financial disclosures — including the detail that its subsidiary Rosetta translates the company's own earnings releases — offer a degree of transparency not available from a private AI developer.
What is MetaReal's geographic focus, and does it have any presence outside Japan?
The company is headquartered in Japan, and its website and disclosed partnerships are predominantly domestic. Its T-4OO engine supports 100 languages, and the firm's mission statement targets a global audience, but no additional office locations or international subsidiaries are identified in current public-facing materials.
Who runs MetaReal — what is the leadership and governance profile?
MetaReal does not publicly profile its management team or board on its corporate website. For a publicly listed entity, this is an atypical posture and leaves an information gap regarding investment decision-makers and governance structure. Institutional allocators typically expect named principals for an entity of this scale.
Does MetaReal engage in venture investing or operate a corporate venture arm?
MetaReal's public-facing materials describe its business as AI product development and enterprise services, not venture capital. Co-development partnerships with Toyo Keizai and Digital Hearts are commercial collaborations rather than equity investments. There is no stated venture allocation or disclosed portfolio of external startup investments.
Profile maintained by Altss using OSINT (open-source intelligence), regulatory filings, licensed data partners, and verified direct submissions. Read the methodology. Last updated: . Continuous refresh with full update cycles at least every 30 days.
Need institutional-grade insight on family offices?
Altss delivers:
Prefer a guided tour?
We’ll walk you through: