The
Zia
Agent
studio
is
now
fully
prompt-based.
Agents
built
by
users
can
be
deployed
autonomously,
triggered
through
button
click
or
rule-based
automation,
or
summoned
within
customer
conversations.
“Zia
LLM
is
a
unique
Enterprise
LLM,
this
investment
furthers
Zoho’s
longstanding
aim
to
build
foundational
technology
focused
on
protection
of
customer
data,
breadth
and
depth
of
capabilities,
and
value,”
says
Gibu
Mathew,
VP
and
GM
of
Zoho
APAC.
“Because
Zoho’s
AI
initiatives
are
developed
internally,
we
are
able
to
provide
customers
with
cutting-edge
tool
sets
without
compromising
data
privacy
and
organizational
flexibility,
democratizing
the
latest
technology
on
a
global
scale.
For
businesses,
AI
needs
data
integration
at
a
DNA
level,
the
Zoho
platform
is
built
with
exactly
that
future
and
vision.”
Zia LLM, Built from the Ground Up and Optimized for Business
Zoho has launched its own large language model, Zia LLM, built completely in-house by leveraging NVIDIA’s AI accelerated computing platform. Trained with Zoho product use cases in mind—ranging from structured data extraction, summarization, RAG, and code generation—Zia LLM is comprised of three models with 1.3 billion, 2.6 billion and 7 billion parameters, each separately trained and optimized for contextual applicability that benchmark competitively against comparable open source models in the market. The three models allow Zoho to always optimize the right model for the right user context, striking the proper balance between power and resource management. This focus on right-sizing the model is an ongoing development strategy for Zoho.
In addition to Zia LLM, Zoho is announcing two proprietary Automatic Speech Recognition (ASR) models for speech-to-text conversion for both English and Hindi. Optimized to perform on a low computer load without compromising on accuracy, the models benchmark up to 75% better than comparable models across standard tests. Language support for additional languages will be coming in the future.
While Zoho supports many LLM integrations for users, including ChatGPT, Llama, and DeepSeek, Zia LLM continues Zoho’s commitment to data privacy by allowing customers to keep their data on Zoho servers, leveraging the latest AI capabilities without sending their data to AI cloud providers.
Zia LLM will be deployed across Zoho’s data centers in the US, India, and Europe. The model is currently testing for internal use cases across Zoho’s broad app portfolio, and will be available for customer use in coming months.
Effective Native AI Agents Ready for Use
To enable immediate adoption of agentic technology, Zoho has developed a roster of AI agents contextually baked right into its products. These agents can be used across various business activities, handling relevant actions based on real-life organizational roles (including sales development, customer support, and account management).Agents available today include:
- New Version of Ask Zia: The latest version of Zoho’s platform-wide conversational AI assistant, Ask Zia’s new BI skills are tailored to data engineers, analysts, and data scientists, yet supports any user within an organization. Ask Zia is now equipped with capabilities that directly address the unique pain points faced by each persona, whether it’s building end-to-end data pipelines for engineers, analyzing data, creating reports and dashboards in an interactive conversation mode for analysts, or helping to jump start building ML models for data scientists.
- Customer Service Agent: With the ability to process incoming customer requests, understand the context behind them, and either answer them directly or triage them to a human rep, the Customer Service Agent for Zoho Desk provides an efficient yet reliable first line of assistance, paving the way for quicker responses and resolutions.
AI Agent Studio and Marketplace
First announced earlier in 2025, Zoho has further simplified the Zia Agent Studio experience to be fully prompt-based (with the option to use low-code) and include ready-made access to over 700 actions across Zoho’s products. Agents built by users can be deployed autonomously, triggered through button click or rule-based automation, or summoned within customer conversations.At the time of deployment, an agent can also be provisioned as a digital employee. Digital Employees respect defined user access permissions, maintaining the same permissions structures already defined within the organization. Admins are able to perform behavioral audits as well as performance and impact analyses on Digital Employees, ensuring that every agent is working as effectively as possible and within clear guardrails.
Zoho Marketplace, which supplies over 2500 reliable extensions and integrations for Zoho users, now houses the Agent Marketplace, a dedicated section for AI agents that can be deployed by customers quickly. Ecosystem partners, ISVs, and individual developers will soon be able to create agents and host them on the Zia Agents Marketplace, further simplifying the adoption of agentic technology by organizations.
Some pre-built agents created with Zia Agent Studio (and available on the Zia Agent Marketplace) are:
- Revenue Growth Specialist: Uncovers opportunities for upsell and cross-sell across existing customers, recommending the best marketing approach for each customer.
- Deal Analyzer: Analyze deals and provide insights such as win probability, next best action, and follow-up suggestions.
- Candidate Screener: Intelligently identifies and ranks the most suitable candidates for a specific job opening based on role requirements, skills, experience, and other key attributes.
With over 55 applications across one ecosystem, users can build agents to meet their organization’s every need, no matter how specific. With Zia Agent Studio, Zoho users have access to the same tools as Zoho’s developers, ensuring that any agent a customer dreams of can be created with ease.
Interoperability with MCP
Zoho has adopted the model context protocol (MCP), offering its own MCP server with a rich action library across several applications, allowing any MCP client to tap into data and actions from various Zoho apps while respecting the customer’s defined permission structures.Zoho’s MCP server has a library of actions from more than 15 Zoho applications exposed during Early Access. With Zoho Flow, third party tools are also exposed. Additional Zoho applications will be onboarded in the coming months. Furthermore, Zoho Analytics now offers support for a local MCP server.
Roadmap
In the short term, Zoho will regularly scale Zia LLM’s model sizes, starting with the first of several planned parameter increases by the end of 2025. Future planned releases include expanding the available languages used by the speech-to-text model, beginning with languages spoken primarily across Europe and India, as well as the introduction of a reasoning language model (RLM).Additional skills will be added to Ask Zia, allowing it to act as an assistant to Finance teams and Customer Support teams, with more skills added in the future.
Support for the Agent2Agent (A2A) protocol will be implemented, allowing for Zia Agents to interact and collaborate with each other, as well as collaborate with agents on other platforms.
Additional Documentation
Zia Agent Marketplace – Full list of available agents at launchDisclaimer
All trademarks, product names, and company names cited herein are the property of their respective owners.Availability and Pricing
Zia LLM will be available to Zoho customers in the coming months. Zia Agents, Zia Agent Studio, Agent Marketplace, and Zoho MCP Server are being rolled out to customers who are currently on the early access waiting list. General availability for these offerings is expected towards the end of 2025. Zoho expects to study the usage patterns of these customers across use cases, industries, geographical regions, and sizes during this early access phase. A pricing structure for these offerings can be expected at the time of general availability.Zoho Artificial Intelligence Differentiation
Zoho is committed to designing and incorporating artificial intelligence guided by the principles of customer privacy and value. Our generic AI models across contextual, assistive, and agentic AI, are not trained on consumer data and do not retain customer information. Zoho builds AI tools with usefulness in mind, striking a balance between providing AI technology that assists workers while right-sizing models that don’t require burdening consumers with additional costs.
https://www.zoho.com
https://www.linkedin.com/company/zohoapac
Hashtag: #zoho
The issuer is solely responsible for the content of this announcement.
Support InfoStride News' Credible Journalism: Only credible journalism can guarantee a fair, accountable and transparent society, including democracy and government. It involves a lot of efforts and money. We need your support. Click here to Donate