[Organizational AI Utilization #233] Gemini Spark is now available for corporate accounts. I have organized the differences between it and the combination of Google Workspace Studio, Gems, and GAS.

Hello! This is Terada.

Currently, I serve as the Representative Director of the AI Digital Community (ADC), a community for AI practitioners such as AI promoters at digital-related companies as well as the Representative Director of FURIKAKE Partners Inc., which supports the "xAI" transformation of client businesses, and AI Portalize Inc., which provides products that support organizational AI utilization, supporting organizational AI utilization from various aspects!

This time, I have organized and summarized the differences between Gemini Spark and conventional tools, following its release for corporate accounts.

*Postscript (2026/08/03 18:42)
It appears that Gemini Spark for corporate use is temporarily disappearing from company screens (as of just past 18:00 on August 3rd). Please be aware that there are cases where Gemini Spark may be unavailable depending on when you check.

Gemini Spark has finally become available for corporate use. Many of you are likely wondering, "How is it different from previous tools?" and "How will our company's operations change?"

To understand Gemini Spark, I believe it is first necessary to know the roles and limitations of the three conventional tools we have used in the past.

After reviewing the prerequisite knowledge, I will explain the features of Gemini Spark and the decisive differences from conventional methods in an easy-to-understand manner!

[Review] The three prerequisite conventional tools and functions

Let's organize in detail the three conventional tools and concepts that are key to understanding Gemini Spark.

1. Google Workspace Studio

This is a no-code platform that allows you to combine various Google Workspace apps (Gmail, Drive, Sheets, etc.) to automate daily routine tasks and workflows.

Even without programming knowledge, its strength lies in the ability to intuitively build linkage processes (triggers and actions) such as "If an email with subject XX arrives, summarize the body and add it to a spreadsheet."

On the other hand, while it is convenient, there were limitations to its functionality. In particular, the fact that "it could only output to one's personal My Drive area and could not directly write to or output to team shared drives or shared files" was a major bottleneck.

*Click here for related articles on Google Workspace Studio

2. Gems

This is a custom AI function that allows you to set and save specific roles or instructions (system prompts) and prerequisite knowledge (reference files) for Gemini, the generative AI, in advance.

You can easily create AI specialized in specific thinking, dialogue, or analysis, such as "an excellent in-house AI researcher" or "a professional proofreading assistant."

However, even though it is good at advanced dialogue and information retrieval, interactions are limited to the chat screen. It could not directly execute concrete practical tasks (Workspace operations or file output) such as "writing search results to a spreadsheet" or "creating a file and saving it to a shared drive."

3. GAS (Google Apps Script)

This is a development environment for automation and extension based on JavaScript that allows you to control various Google Workspace tools and integrate with external systems and APIs.

It allows you to build highly advanced and flexible automation systems, not just within Google Workspace, but also by connecting to APIs of external web services or sending automatic notifications to external tools like Slack. Naturally, it also allows for free access and manipulation of shared drives and shared files.

However, while it is extremely powerful, it requires specialized knowledge (IT skills) to write programming code. Because it involves the effort of writing code, maintenance, and error handling, it has been a high hurdle for non-engineers.

*Click here for past articles related to GAS

By combining these three tools, it becomes possible to realize even relatively complex workflows.

Also, depending on the situation, some tasks can be completed solely with Google Workspace Studio, meaning there are cases where you do not need to connect Gemini or use GAS.

However, by combining these three, you can do a wider and more diverse range of things.

Main features and points of caution for Gemini Spark

Gemini Spark is an AI agent that operates autonomously just by giving it instructions.

Creating these three tools also becomes easier by utilizing AI. It has become easier to build and change workflows, and some of the original constraints have been relaxed.

With that in mind, I have organized three points that are particularly important to look at.

1. Web search and browser operation

As a feature, it can refer to specific websites by directly specifying a URL, and it can also explore and search the web on its own to collect necessary information without a specified URL.

As a constraint, it cannot perform browser operations on sites that require a login. Even if you are logged in on your PC browser, Gemini Spark executes processes in the backend system, so it cannot load screens that appear after logging in.

2. Google Workspace file and data operation

As a feature, it can operate across tools within Workspace, such as creating email drafts in Gmail or registering and retrieving appointments in Google Calendar, in addition to file operations within Google Drive (creating, moving, copying, and operating shared drives).

As a constraint, it cannot directly operate local files. It cannot directly read or write files saved on your PC, so you must first upload them to Google Drive or similar. Also, there are currently unsupported operations and functional limitations for each tool.

3. Integration with external services

As a feature, it can flexibly integrate with various external services in the form of "apps," not just within Google services.

As a note of caution, there are parts of external integration that have not been fully tested yet, so it is necessary to verify in advance whether all integrations will work without issues.

The decisive difference from conventional tools

Previously, for example, if you wanted to incorporate Gemini into Google Workspace Studio and link it with external services, you had to write code in GAS to combine them.

Gemini Spark is a tool that allows you to accomplish tasks that previously required multiple tools all at once, simply by giving instructions in natural language (Japanese).

These are the points of evolution

The biggest advantage is that the learning cost of mastering new tools like Google Workspace Studio has been drastically reduced. By utilizing AI, it has become easier and faster to create workflows than ever before.

As for GAS, it was already becoming easier to write by using AI. The position here is that the construction work itself has become even easier than that.

Also, the 'output (writing) to team shared areas,' which was a major hurdle for no-code automation, can now be easily overcome.

With the simple Google Workspace Studio alone, you could only output to your 'My Drive' area, and to save to a shared drive, you had to combine GAS code to create a workaround. However, by using Gemini Spark, you can bypass the hassle of writing such GAS code and the functional limitations of Google Workspace Studio all at once.

There are two major changes

The first is the reduction in tool learning and configuration costs. Not only do you not have to learn GAS code writing or Google Workspace Studio screen operations from scratch, but the 'stage of thinking about means and processes,' such as what to connect and in what order, has become unnecessary just by giving instructions in text.

The second is direct access to shared areas. 'Direct output to team-shared files and folders,' which was difficult with Google Workspace Studio alone, can now be easily specified and executed just by using natural language instructions without writing GAS.

[Use Case Comparison] Sharing Google Meet recordings and transcripts

As a concrete example, let's compare the 'conventional way' and the 'Gemini Spark way' when automating the following business task.

The target task is 'continuously saving data recorded and transcribed in Google Meet to a shared drive and automatically transcribing summaries of that content to a shared spreadsheet.'

Conventional way (combination of Google Workspace Studio and GAS)

Since you cannot output to a shared area with the no-code Google Workspace Studio alone, you had to combine GAS and write code.

• Step 1: After recording a Meet, copy the transcript saved in My Drive to a shared drive (GAS, requires code writing)

• Step 2: Triggered by the file in My Drive, create a summary (Google Workspace Studio)

• Step 3: Transcribe the created summary to a spreadsheet under your own authority (My Drive) first (Google Workspace Studio)

• Step 4: Reflect the transcribed content in the shared drive spreadsheet (GAS, requires coding)

How to do it with Gemini Spark

There is absolutely no need for coding, cross-tool processing, or workarounds.

• You can handle the entire process end-to-end using only Gemini Spark

• All instructions are completed with Japanese prompts, and no knowledge of GAS is required

• Once you register a series of processing rules as a "skill," you can automatically reproduce them from the next time onwards

An era where shared task automation can be done with just Japanese instructions

With the lifting of corporate usage for Gemini Spark, we have entered an era where "automation of team sharing tasks" can be realized by anyone simply by giving instructions in Japanese.

Until now, specialized skills such as GAS programming were required, but this can now be covered by AI. This lowers the psychological hurdle for construction and makes it easier to realize complex configurations that previously required combining multiple tools.

In particular, I think it is a good fit for those who have the following challenges:

• Want to automate tasks without the cost of learning programming or new tools

• Want to leave tasks that span tools within Workspace (Gmail, Calendar, Drive, etc.) to AI

• Want to smoothly automate information sharing within a team using shared drives

Please try starting by choosing one routine task you do regularly and asking Gemini Spark to do it in Japanese!

Thank you for reading!
Click here for past AI-related articles!

By the way, I have also published a book summarizing AI promotion in organizations. Please take a look if you like.