First time login (registration)

When you access Maison AI , the following login screen will be displayed.

There are two ways to register for the first time:

A: Log in using your Gmail account via Sign in with Google

B: Register using your email address

If you register using an email address, please be sure to use the same web browser from registration to receiving the email and then moving to the registration page.

Create workspace

Once you have successfully logged in, if you have not yet opened a workspace, the following screen will be displayed. Please create a new workspace from here.

Create an account

Enter a name for the workspace.

The name you choose here can be changed later.

Create a workspace

Enter a name for the workspace.

Invite members

Invite members to join the workspace.

You can also skip this screen.

Enter the email address of the member you want to add and click "Send Invite."

When you invite a member, an invitation email will be sent to the email address you entered.

Once you have completed the above steps, you will be redirected to the workspace home page.

If you already have an account

Inviting workspaces

On the workspace selection screen, the workspace you've been invited to will be displayed in the "Invited" column.

Enter the name you want to use for the workspace

If you don’t have an account yet

Create an account

Enter your email address to create a new account

If you are successfully logged in for the first time, the following screen will be displayed.

Enter a name for the workspace.

The name you choose here can be changed later.

Select a workspace

The above screen will be displayed the second time you log in or later.

*This will not be displayed the first time you log in.

Participating

A list of workspaces that you can join is displayed.

Invitation

This will display a list of workspaces you're currently invited to.

If you don't see the workspace you want to join, you can get an invite from someone who's already in it to join your account.

overview

This page explains the [Workspace Settings].

[Workspace Settings] is only visible to users with Owner/Administrator privileges on the Basic plan.

What is workspace registration?

After the trial period ends, you will need to register your credit card and make a paid registration to use MaisonAI's workspace.

Please see the plan page for details on fees and plans. Only users with owner privileges can register/cancel.

Details of each permission can be found on the Account Permissions page.

Register your workspace

This can be done by going to [Workspace Settings] → [Start Full Registration].
*Only users with owner privileges will see the "Start registration" button.

The cost of this registration varies depending on the number of members participating in the workspace. Please remove any members you do not need in advance.

Adding/removing individual members

After you register your workspace, the cost of adding members will be calculated on a pro rata basis. There is no cost when you invite someone, but it is charged when they join.

Both owners and administrators can add members, but only the owner can view payment history. Administrators should make adjustments and add users to avoid unexpected charges.

There are no pro-rata refunds for removals, but if a removed member rejoins before their renewal date there will be no additional charge.

For instructions on how to add/remove members, please see the Managing Members (Add/Remove) page.

Rename a workspace

Users with Owner privileges can change the workspace name in the field below.

overview

This page explains how to cancel or change your credit card information in Workspace Settings.

Cancellation and changes to credit card information can only be made by users with Owner privileges for the Basic Plan.

Cancel a workspace

You can do this from Workspace Settings → Cancel Workspace.

When you click [Cancel workspace], you will be redirected to the [Stripe history page], where you can select [Cancel plan].

Even after cancellation, you can still use the service until your next renewal date.

After cancellation, your workspace data (history, agents, etc.) will be stored for a certain period of time and then deleted.

Change your credit card details

• As an Owner, go to Workspace Settings.
• Click Check Payment History.

• Click Add Payment Method and enter your new card details.

• Make the new card the default.

overview

On the Member Management Screen, you can see who is already in the workspace and who is in the process of being invited.

After you register, the price will vary depending on the number of members. Please adjust the number of members you do not need in advance.

To check the permissions of all participating members, or to delete members, or to delete invited members (cancel invitations), click the Members item on the right side of the home screen header.

① Narrow down the members

• You can narrow down your search by user name or role.

② Inviting members

• You can invite members to your workspace using the invite button.
• You can grant permissions when inviting someone.
• If you invite a user who has already used Maison AI (a user who has joined another workspace), they will be displayed in the member list rather than being invited.
• There will be a cost when an invited user joins your workspace.

③ Deleting a member

• Users with Admin/Owner permissions can remove other members from the workspace.
• Any user can remove other invited members from a workspace.

④ Change member permissions

• Users with admin /owner privileges can change the privileges of other members.
• Ownership cannot be changed.

Member management screen on smartphone

On the smartphone version, access the member management page from the left menu.

Maison AI has limited functionality depending on the account permissions you are given.

Restrictions on functions based on permissions

Changes to permissions

You can change it by selecting from the drop-down menu on the member list page .

* Only Administrator → General / General → Administrator. Owner authority cannot be changed.

How to use

①Select workspace

• The workspaces you are a part of are displayed.
• You can switch workspaces by selecting them.

②Each function

You will be redirected to each page.

The page you are currently viewing is highlighted.

• Get instant access to useful features such as home learning content, AI Agent Pro, and AI design templates.

• You can use the original ChatGPT without configuring AI chat behavior.

• AI Agent Pro
  You can use AI Agent Pro to chat immediately without any setup.

• AI Agents
  You can view the list of AI agents and use AI agents.

• AI templates
  You can view the template list and use the templates.

• AI Design
  You can use image generation AI.

• Prompt Library
  You can use AI functions set for specific purposes such as general purpose, fashion-specific, and image.

• history
  You can view the chat history.

③Workspace settings

④ Account settings

overview

You can access various functions from the workspace home screen.

Header

• At the top of the page is the header.
• By entering your question here, you can start a regular AI chat without an agent.
• Go to the member management page.

Learning Content

• You can watch the MaisonAI tutorial video.
• Click on a thumbnail that interests you in [Recommended Learning Content] and the video will begin playing.

AI Agent Pro

• We have prepared AI Agent Pro, which even first-time users can use easily.
• It can be used directly from the home screen, but unlike regular AI agents, it cannot be customized.

AI Design

• You can generate images.
• By clicking on each sample, you can start generating images similar to the sample, just like with AI Agent Pro.

overview

This feature provides task-specific prompts for those who want to quickly and easily create AI agents and AI templates.

The content provided includes templates, AI agents, and prompts for image generation.

The library of content will continue to grow, and users can use these prompts as is or customize them to suit their needs.

Please note that there is no functionality for users to post their own prompts.

How to use

① List screen

• Access the Prompt Library from the left menu.

• There are two types of prompts that generate text: common and fashion.

• There is one type of prompt that generates an image.

②Details screen

• On the [Details] screen, you can check the content of the prompt and the situations in which it is used.

• Clicking [Create] will take you to the [Editing screen].

Editing screen

• Clicking [Create] will take you to the [Editing screen].

• On the [Edit screen] you can customize the prompt content and name .

• Click Create to add it as an agent/template.

Use Case (Image Prompt)

① Image generation using the reference image

• Clicking on [Image Generation] will take you to [AI Design].

②Customize your images

• By rewriting parts of the prompt, you can customize it to suit your needs .

overview

In [History], you can view your own history and that of other team members. This can help you get ideas for how to use AI, and by sharing your own usage with others, you can promote the use of AI within your team.

History list screen [Display]

How to use

①Text/Image Switching Tab

This tab switches between the following two displays:

• AI chat, AI agents, AI template generated text
• AI-designed images

②User switching tab

This tab allows you to switch between the following three displays:

• all
• View the history of all users.
• member
• View the history of members other than yourself.
• You cannot see history that members have set to "private."
• myself
• View your history.

③ Search

You can search not only the title of the history but also the content of the conversation.

④New chat from history template

You can now go to the template you used and start a new chat.

History list screen [Edit / Delete]

On the list screen, you can make various edits to your own history.

How to use

• Select [Edit] from the [...] icon. You can also [Delete] it.

• You can set easy-to-understand titles for your history and change the scope of visibility.

History details screen

① Details of past history and resume chat

You can also [Go to used template] from the details screen.

overview

Your account settings allow you to customize the experience that works best for you.

You can switch the language (Japanese or English), AI model (GPT-3.5 or GPT-4), and change your username.

Feel free to adjust these settings as they won't affect other members.

How to use

• Click on the account name in the left menu to go to the account editing screen.

① Change the icon

• You can set the icon image.
• By default it will be your initials.

② Change your account name

• You can change the account name.

③-④ Change your email address and password

• You can change the email address and password you use to log in.
• This will not be displayed if you created your account with Google Login.

⑤Changes in the text generation model

• The accuracy of text generation AI such as AI agents varies depending on the AI ​​model used.
• "ChatGPT3.5" is set as the default.

⑥Change the image generation model

• The functionality and accuracy of image generation will vary depending on the AI ​​model used.
• "Stable Diffusion XL" is set as the default.

Notes

• The upper limit varies depending on the AI ​​model used.
• Please refer to the pricing page for the maximum limits for each model.

⑦Language

• MaisonAI can be switched to four languages. The following are supported:
• Japanese, English, Chinese (Traditional), Korean

Notes

• Even if you switch the language, if a prompt for an AI agent or other device specifies "Please output in Japanese," the AI's response will be in Japanese.

Users can use AI agents and templates in combination with text generation AI technologies such as ChatGPT to engage in dialogue. The content generated here can be used to improve work efficiency and plan new businesses.

Specifically, you can elicit new ideas through question-and-answer format through dialogue with an AI agent, or use the template function to generate standardized answers.

Chat Basics

There are four types of chat

Selecting a Text AI Model

• Standard model: ChatGPT 4o-mini
  • The response time is fast, so you can chat without waiting.
  • You can use it without worrying about the upper limit.
• High performance model: ChatGPT-4o
  • It gives a more logical and human-like answer.
  • There is a monthly usage limit per workspace.

Upper limit of text AI model

• Maximum number of characters that can be entered per question: 40,000 characters
• Maximum total number of characters per chat: Approximately 96,000 characters

How to use

• Go to Account Settings.
• Select [Model to use] > [Model name] .
• Press [Change].

overview

This page compares user-customizable AI agents and AI templates, explaining the purpose and mechanisms of each.

For information on AI Agent Pro, which cannot be customized by users, please see here. AI Agent Pro

Purpose of AI agents and AI templates

①AI Agent

You can set the AI's behavior in advance before using it.

Settings include the AI's professional background, personality and tone of voice, accumulated knowledge, and the format of the text to be output.
This makes it possible to recreate an "AI expert" that is knowledgeable in a specified field and behaves as specified.

AI agents are good at a variety of tasks, from simple to abstract and expert advice.

example:

This is an AI agent that creates product descriptions from bullet points.
The purpose and output format are specified on the AI ​​agent setting screen.

②AI Templates

The first sentence you send to the AI ​​is set as a "template" and used.

When starting a chat, users can fill in the blanks or customize the text.
This allows you to send complete information to the AI, making it easier to get targeted answers.

This is great for simple tasks where you want repetitive or standard output.

example:

This is an AI template that creates a catchy slogan for a product.

overview

Since the list screens of AI agents and AI templates have many items in common, we will explain the AI ​​agent list screen in detail here.

On the list screen, you can easily select the agent you want to use and access the chat screen. Various functions are provided to help you smoothly find the agent you are looking for from the many agents available.

About each function in the list

① Team/Private tab

You can switch agents depending on the scope of visibility.

Agents on the team can use it too, and chat history is shared across the team.

Since private agents can only be used by you, you can use them for tasks that are specific to your work or that do not need to be shown to others.

You can change the visibility of your chat after it has been posted.

How to use

• You can switch between them by clicking the tabs.

② Search

You can search by agent name or text in the description.

How to use

• Enter the word you want to search in the search box.
• AND searches are also possible.
• Search results are updated in real time.

③ Three dots

Clicking the [...] icon allows you to change the settings of that agent on the list screen.

How to use

• Click the [...] icon.
• Templates that users have created themselves or have administrator privileges or higher can be edited or deleted.

④ Favorites

You can register frequently used agents as favorites.

The Favorites feature is not shared with other members, so each user can favorite things individually.

How to use

• Click the [★] icon to the right of the agent you want to make your favorite.
• The selected agent will now be displayed in [Favorites] at the top of the list screen.

⑤Add

Add a new agent. Clicking this will take you to the agent settings screen.
For details, please see the AI ​​Agent/AI Template settings screen here.

overview

This section explains the settings screens for the AI ​​agent and AI template.

As with the list screen, there are common items between AI agents and AI templates in the settings screen. In addition to these common items, we will also explain the unique items that each has.

Common items

Common 1) Title, description, and icon image

The settings you make here will be displayed on the list screen.

title

• The name of the agent. It does not affect the AI ​​generation results.

Description

• This is a description of the agent. It does not affect the AI ​​generation.

Icon Image

• You can attach an image to your AI agent.
• Please upload a square image.
• The default is no image.

Common 2. File upload

• We will answer based on the file information uploaded here.
• You can attach PDF, CSV, Word, and text files.
• The character limit varies depending on the GPT model you use.
• Accuracy will improve if you write "Please answer according to the attached file" in the agent settings.

Common 3) Scope of disclosure

team

• Publish the AI ​​agent to members of your workspace.
• Chats created in a team will have their history visible to the team.

Private

• Your AI agent will not be visible to other members of the workspace.
• Any chats created privately will have their history visible only privately.

You can also change the visibility of your chat history after the chat.

Common ④Create Button

• Create an agent with the configured settings.
• The user or administrator who created it can edit the contents later.

AI Agent Settings

What kind of AI agent is it?

Set the behavior of the AI ​​agent. For example, there are the following items.

• Areas of expertise and prior knowledge:
• Set specific areas of expertise and prerequisite knowledge that the AI ​​agent should have, such as "marketing expert," "technical support specialist," "knowledge of SEO," and "customer support process."
• Personality and tone:
  • Customize the personality and tone of your AI agent. For example, you can set it to be "friendly" or "formal."
• Output format settings:
  • Set the format and style of the text generated by the agent, for example "bullet points" or "detailed description".

First message

Enter a message to start the chat, giving clear instructions to users on how to start the chat and what the AI ​​agent can answer.

Example of message to set:

• "In this chat you can ask questions about product information."
• "This AI agent will provide advice on marketing strategies."

These messages provide users with a quick guide to get the information they need.

AI template settings

Templates

With AI templates, users can always get answers with a consistent level of accuracy by rewriting or filling in parts of pre-set template sentences.

When creating a template, it is a good idea to consider the following:

① Output instructions to AI:

Clearly define the output you want from the AI. For example, enter specific requests such as the following:

• For example: Please briefly explain the features of your product.
• Example: Please respond appropriately to the inquiry.

② Input field for users:

To use AI templates effectively, it’s a good idea to provide clear input fields for users so they know what to fill in. We recommend including specific instructions in your template, such as:

• Example: "Enter product name here: _____"
• Example: "Please enter your inquiry: _____"

overview

AI Agent Pro is an AI agent that does not require any initial setup and can be tried right from the home screen. Even first-time users can easily use it.

However, unlike normal AI agents, users cannot customize the AI ​​settings.

You can get advice from expert agents in a wide range of professions.

How to use

overview

Image analysis can be performed on the chat screen. This allows AI to analyze the characteristics of uploaded images.

Specific uses will be introduced as "use cases" later in this article.

How to use

• On the chat screen, select the image from the paperclip icon.

• After selecting the image, the analysis is performed along with the text that instructs it. Note that multiple images can be analyzed at the same time.

Notes

• Image analysis supports the following file extensions: PNG, JPG, JPEG, WebP, GIF.
• Image file size is limited to 20MB

Use Cases

We will introduce some use cases that may be useful for image analysis.

① Analyze images and create product descriptions

The AI ​​writes the introduction by analyzing the characteristics of the image and specifying the length of the text etc. in the instructions.

② Analyze the image and create a prompt for image generation

From image analysis, you can also use text descriptions of images to generate similar images. Below, we instruct the AI ​​to analyze by category, such as [structure], [shape], etc.

Image generation results

3) Analyze handwritten notes and diagrams

You can also photograph and analyze handwritten notes and diagrams, and use AI to transcribe them and analyze the diagrams.

overview

Users can upload files such as CSV and Excel on the chat screen, making it easy to analyze constantly changing data such as sales and reports.

How to use

Model Switching

File uploads can be performed using both GPT-4o mini and GPT-4o.

There are differences in accuracy, so choose which model to use based on your needs.

Uploading via chat screen

On the chat screen, you can select the file you want to upload using the [+] button next to the send button.

You can also select multiple files.

Send it to chat with instructions on how to analyze it.

Available file formats

The supported file extensions are as follows:

For more information about image analysis, please see below.

Link: Image Analysis

Upper limit

The upper limit for each model is as follows. Both GPT-4o and 4o mini can handle the same number of characters.

Please note that this limit includes not only the number of files uploaded, but also the cumulative number of chat characters. If you reach the limit, please start a new chat and continue using the service.

Use Cases

It can refer to multiple CSV files to suggest appropriate actions and strategies.

As a specific example, we refer to product inventory and sales data to ask which products need promotion and what promotional techniques would be effective.

The AI ​​agent can analyze these CSVs and make suggestions, including promotional methods.

It is also possible to perform analysis with various combinations, such as images and CSV files.

overview

You can now access Google searches and specific websites directly from the chat screen, allowing you to get the latest news and information on the Internet, as well as summaries of websites you specify.

External site references will always use GPT-3.5 regardless of user settings.

Also, please note that when retrieving website content, tokens will be consumed according to the amount of text .

How to use

External reference ON

Every chat screen has a switch to turn external references on or off.

If you want to use external references, turn it ON before starting a conversation.

Once the conversation has started, you will no longer be able to switch it on/off.

If the user does not specify a URL

MaisonAI will perform a Google search to provide you with an answer.

If you simply ask a question, the AI ​​may answer without searching the Internet. If you want it to search and answer your question, try the following phrases:

• Words related to real-time and internet, such as "latest" and "online"
• Instructions such as "Please search on Google and then answer"

Also, if you want to clarify which website you actually looked at when answering a question, it's a good idea to include phrases like the following.

• Instructions such as "Please provide the URL of the quote"

When the user specifies a URL

The AI ​​can refer to reference URLs by pasting them into the conversation.

remarks

You can determine whether AI is using the external site reference function by looking at the display below.

• (Looking up information...)
• If this display is not present, the AI ​​will answer from its own database rather than from an external site.

overview

It can generate images from keyword or text prompts.

This page explains the AI ​​design screen and how to operate it.

How to use (start generation)

① Prompt

• Please enter prompt.
• The information entered early in the prompt will be generated first.
• It can also be generated in Japanese.

② Negative prompts

• Enter what you don't want in the image.
• This input is optional, and you can generate it without any problems.

③Image Style

• There are 18 different image styles to choose from.
• The choice is optional.
• For more information, please see here .

④Image size

• You can choose from 9 different image sizes.
• The choice is arbitrary, the default is 1024x1024 (square).
• For more information, please see here .

⑤Number of copies generated

• You can choose how many images you want to generate with the prompts you enter.
• The minimum number is 1 and the maximum is 10.
• The choice is optional, the default is four, but the more you choose the longer it will take to generate.

⑥ Scope of disclosure

• You can specify whether the information is visible to the team or private.
• Once generation has begun, it will no longer be possible to change the scope of availability.
• You can resume image generation by going to the page of the AI ​​design you created from the history.

⑦New Image

• Clear the images you have created so far
• You can view the cleared history from the " History " menu bar on the left.

⑧ Execution

• Click Run to start the generation.

⑨Automatic translation

• It is set to ON by default. Please set it to ON if you want to input Japanese at the prompt.
• For more information, please see here .

Please note:

The generated images will be deleted after one month, so please download them as needed.

How to use (after generation)

①Restart

• Start generation again while keeping the original prompt information

② Larger display

• Hover the cursor over the generated image to display
• Enlarge the selected image

③Additional corrections

• Hover the cursor over the generated image to display
• Prompts allow you to add any elements you want to the selected image
• Regenerate the image by including the new prompt element in the original image

④Image information

• Displays the information used to generate the image.

⑤Download

• Download the generated image

overview

Starting in June 2024, we have added Stable Diffusion3 to the models that can be used in AI design.

Specific benefits will be introduced in the "Use Cases" section later in this article.

Please note that the image limit for the previous Stable DiffusionXL and the newly added Stable Diffusion3 are each counted separately.

Please refer to the pricing page for the maximum limits for each model.

How to use

① Model switching

• From [Account Settings], select [Model to use] > [Stable Diffusion3] and click [Change].
  *It is also possible to revert to the previous [Stable Diffusion XL].

②AI-designed UI

• The previous Stable Diffusion XL [Image Style] is no longer available. From now on, please specify the expression method such as [Line Art] or [Pixel Art] directly in the prompt.
• Instead of [Image Size] it is now [Aspect Ratio].
• You can now select the output format.

Use Cases

①Improved prompt reflection accuracy

The text you type is now more accurately reflected in the image, especially when it comes to detailed instructions about the number of people in the image and their clothing.

prompt:

Design, entertainment, democratization of reality, various perspectives and uses, high fashion editorial photography, background color is white . The three women are wearing blue, white and red outfits respectively ( left woman: blue, middle woman: white, right woman: red )

②Improvement of character generation accuracy

Please note that it is not possible to generate characters other than the alphabet, such as hiragana or kanji, within an image.

prompt:

[Logo name] "Stylish Shirts Co."
[Design style] Modern and simple [Font] Sans serif, bold [Color] Deep blue and white [Icon] Small silhouette of a shirt [Layout] Icon on top, logo on bottom, surrounded by a circle [Background] White

Stable Diffusion XL

3) Improved accuracy of additional image correction

It is now possible to adjust the strength of the generated image when it is corrected again by AI. This makes it possible to make effective adjustments while maintaining the composition and pose to a certain extent.

*This is a function to edit the generated image. Currently, there is no function for users to freely upload images.

Original prompt:

Line art, cat, full body

Additional correction prompts:

Cat, realistic

overview

Users can brush over parts of the generated AI image and regenerate only those areas, making it easy to fine-tune image details or modify specific areas.

You can also upload any image and have it generated by AI. For example, you can use AI to process photos you have taken yourself.

Please note that these features are only available in Stable Diffusion3.

How to use

① Model switching

• From [Account Settings], select [Model to use] > [Stable Diffusion3] and click [Change].

② [Edit selected area]

• When you hover the cursor over the generated image, a menu will appear, so click [Modify Selection].

③ Generate from uploaded image

• Click [① Generate from uploaded image] from the top panel.
• Click [②Select File] and select the image.
• After that, fill in the area and generate it in the same way as before.

Tips

For more detailed tips on how to use it, please see the page below.

Link: Image generation of the selected area Prompt tips

Use Cases

Design Changes

You can specify areas using product images taken by users or AI-generated images and make design changes to them.

Face Swap

By looking at images of people wearing the product, you can compare different races, age groups, and atmospheres.

*Faces generated from multiple images will not be of the same person.

Change your outfit

You can try out various outfits by looking at images of the product being worn .

Change background

By specifying only the background part, you can try out various situations.

overview

By turning on "Auto-translate", you can automatically translate image generation prompts from multiple languages ​​into English.

How to use

• "AI Design" has automatic translation turned on by default. It is possible to generate it in Japanese as is.
• Please turn this off if you want to prevent unintended changes caused by machine translation.

• Try putting in a variety of words.

• One thing to note is that if you use hiragana, it may be automatically translated into unintended English words.

• In this case, the desired image is generated by converting "hito de" to "hito de".

• It also supports input of Chinese (Traditional), Korean, and other languages.

For details on each image size, please see below.