エクサベース AI、Google最新「Gemini 3.6 Flash」提供開始

Exa Enterprise AI, a subsidiary of ExaWizards (a publicly listed AI company headquartered in Minato Ward, Tokyo), announced on July 21 that exaBase AI—its flagship corporate generative AI service—will now offer Google's latest model, Gemini 3.6 Flash. The body describes exaBase AI as holding the #1 market share in corporate generative AI services, according to Fuji Chimera Research Institute's "2026 Generative AI/AI Agent Market Overview" (2024 results, third-party dialogue-type generative AI application vendor share).

Gemini 3.6 Flash is positioned as a flagship model suited to coding, knowledge work, and multimodal processing. It delivers two primary improvements over its predecessor, Gemini 3.5 Flash. First, coding performance is enhanced: developers generate more accurate code with fewer unintended modifications and rework cycles. Second, token consumption is reduced—the same work can be completed with fewer output tokens, delivering faster processing and improved cost efficiency. Additionally, the body notes that safety measures have been strengthened against high-risk use cases in chemistry, biology, radiation, nuclear (CBRN), and cybersecurity domains.

Existing exaBase AI users gain immediate access to Gemini 3.6 Flash without additional application or special conditions. The platform uses the same interface and workflow, allowing users to experience the new model's features right away. exaBase AI has served approximately 1,700 companies since its paid service began in June 2023. The platform is designed with corporate security and compliance in mind, supporting admin visibility into usage, prohibited-word registration, and the ability for users and administrators to upload proprietary files and conduct dialogue and generation based on that company-specific data.

exaBase AI has established itself as the market leader in corporate generative AI services, and the integration of Google's Gemini 3.6 Flash represents a significant technical upgrade for its user base. The new model addresses two core pain points for enterprise users: coding quality and cost efficiency through token reduction. The body notes that Gemini 3.6 Flash is positioned as a flagship model suited to coding, knowledge work, and multimodal tasks—the breadth of use cases that define current enterprise AI demand.

For companies already invested in exaBase AI, the upgrade is frictionless; the platform removes technical and administrative barriers to adoption. The emphasis on safety measures (particularly in CBRN and cybersecurity domains) signals that exaBase AI and Google are directly addressing regulatory and risk-management concerns that corporate buyers weigh heavily when selecting AI tools. This is consistent with the body's description of exaBase AI's existing compliance and security features, which include admin oversight, prohibited-word registration, and support for company-specific data uploads.

No comments yet. Be the first to share your thoughts!