AIUNIFY PROS SUITE can work with multiple AI models and provider families. When model selection is available to your account, the AI Model Selector appears in the Header and allows you to choose which available model will be used by supported AI workflows.
Model selection can affect:
The basic workflow is:
Open AI Model Selector
↓
Review Available Models
↓
Choose Model
↓
If available:
Model Selected
or, if credentials are required:
Model Not Available → Settings → Configure API Key → Select Model Again
↓
Review Tool Token Cost
↓
Generate
The current available Model list is loaded dynamically, so this manual does not treat one fixed list of model names as permanently guaranteed.
When AI Model controls are available, the selector appears in the main PROS SUITE Header.
The control displays either:
or:
if no recognized current model is displayed.
Selecting the control opens:
Select AI Model.
The Header only displays the AI Model Selector when AI settings are enabled for the current PROS SUITE experience.
Therefore, if you do not see a Model Selector:
This manual covers only the subscriber controls that actually appear.
To review available AI Models:
When PROS SUITE is still retrieving the available Model collection, the dropdown displays:
Loading models...
Allow loading to complete before choosing a Model.
The list can be refreshed when provider credentials change.
One of the most important concepts in this chapter is:
The Model dropdown is the authoritative user-facing list of Models currently available to you.
PROS SUITE begins with a source-defined Model catalogue but also requests an updated Model collection from its AI Model service. If that request returns Models, the application updates the available list.
Because of this, Model availability can change without requiring this User Manual to be rewritten every time.
AI Models can change over time.
A Model that appears in:
may not necessarily be available in your current Model Selector.
Always use:
Header → Select AI Model
to determine what is currently available.
Each Model entry in the Selector can display:
The current Model appears with a small indicator on the right side of its entry.
This makes it possible to see both the identity and intended character of the Model before selecting it.
Model entries include short descriptions intended to help distinguish their general characteristics.
The source-defined Model catalogue contains descriptions such as:
Use the description as a quick guide rather than assuming every Model is intended for exactly the same type of work.
The current PROS SUITE model architecture recognizes Model provider families including:
The specific Models available under each provider can change.
The current source contains a fallback catalogue that includes Models associated with providers such as NVIDIA, Google, OpenAI, Anthropic, Mistral, and Groq.
For example, the source-defined collection contains model families such as:
However:
This is not a promise that every one of those Models will always appear in your live dropdown.
Use the current Model Selector as the operational reference.
The source-defined default Model ID is currently:
Gemini 3.6 Flash.
When no saved user Model preference exists, PROS SUITE initializes Model selection using its default Model definition.
Although the source defines a default Model, the available Model collection itself is dynamic.
If the current Model collection changes and the automatically selected Model is no longer valid, PROS SUITE can return to its configured default when the user has not established a manual selection.
For users, the simplest rule is:
Look at the Model Name shown in the Header to determine your current Model.
When you manually choose a Model, PROS SUITE stores that selection in the browser.
This allows the browser to remember your preferred Model between sessions.
Because your selected Model is stored locally in the browser, the selection should not be assumed to automatically synchronize to every device.
For example:
Desktop Browser
→ Model A
can coexist with:
Laptop Browser
→ Model B
You may need to choose your preferred Model again on another browser or device.
If browser site data is cleared, the saved Model selection can be removed.
When no saved selection exists, PROS SUITE returns to its current default Model behavior.
If your Model unexpectedly changes:
To change the active Model:
Changing the Model only changes the selected AI Model for supported workflows.
It does not automatically:
After selecting the Model, return to your Tool and initiate the AI operation normally.
PROS SUITE can recognize that a Model already has the required access configuration available.
When the necessary configuration is available, selecting the Model simply activates it.
You do not need to enter a personal API Key simply because a Model belongs to another provider if PROS SUITE already considers that Model configured and available.
If PROS SUITE does not find the required provider access for the Model, selecting it can open Settings and display:
Model Not Available
with:
“This model requires an API key. Please configure it in Settings.”
This does not mean the Model Selector itself is broken.
It means the selected Model currently requires provider credentials that are not available to the user session.
If you receive:
Model Not Available
Chapter 11 covers API Key configuration in detail.
When a Model requires an API Key that is not currently available, PROS SUITE automatically opens the Settings dialog as part of the unavailable-Model workflow.
You do not need to manually search for Settings after the warning appears.
If a Model belongs to OpenAI, configuring an unrelated Anthropic Key does not satisfy the OpenAI provider requirement.
Likewise:
The Settings interface provides separate credential fields for these provider families.
The current subscriber Settings interface contains:
You do not need to configure all of them.
Configure only providers you intend to use and for which your PROS SUITE workflow requires a user-supplied Key.
When API Key information changes, PROS SUITE refreshes the available Model collection.
Therefore, after adding a provider Key:
The current Header selector contains a special case for a Model with the ID:
gemini-2.5-flash
If that Model appears in the available list, the Header allows it to be selected directly without performing the same API-Key availability check applied to other Model entries.
This should not be confused with the source-defined application default, which is currently Gemini 3.6 Flash.
Because Model availability is dynamic, simply use the current dropdown behavior presented to you.
The current source contains multiple Gemini-related references, including different Model versions and the separate selector handling described above.
This is another reason this manual does not tell subscribers:
“Always select Model X.”
Instead:
Use the available Model list and current PROS SUITE defaults as presented in your account.
PROS SUITE maintains a shared selected Model state that is used by supported AI workflows.
For example, the common streaming process sends the current selected Model along with the Tool request.
Specialized applications can also use the same Model choice for their own AI operations.
PROS SUITE contains different kinds of applications:
Although many workflows use the shared selected Model, this does not establish that every Tool uses every Model in exactly the same way.
A Tool may:
Follow the dedicated Tool chapter when specific Model behavior matters.
The standard ToolPage sends the currently selected Model into the shared AI generation workflow.
The shared generation request includes:
This makes the Header Model selection relevant to many of the Suite's standard AI generators.
As established in Chapter 9, the shared standard ToolPage checks the selected Model's provider when calculating the displayed Token cost.
The current rule is:
Base Cost
for the normal standard calculation
and, when the selected Model is recognized as an OpenAI-provider Model:
Base Cost × 1.5, rounded upward.
Whenever you change Models inside a standard ToolPage workflow:
This is especially important when switching between provider families.
The current 1.5× rule is verified for the shared standard ToolPage.
It is not established as one universal formula for every specialized PROS SUITE operation.
For example:
can have their own usage behavior.
Do not manually calculate their costs using the standard ToolPage rule unless that workspace itself establishes the same behavior.
A Tool defines the task.
A Model defines the AI engine used by supported workflows.
For example:
Content Writer
= Tool
Selected AI Model
= Model used to process that supported writing request
Both decisions matter.
For most work, begin by selecting the correct specialized Tool.
For example:
Need an Article
→ Article Writer
Need Google Ads
→ Google Ads
Need a Business Plan
→ Business Plan
Then consider whether a different Model is necessary.
Do not use Model selection as a substitute for choosing the right Tool.
Consider changing Models when:
Leave the current Model selected when:
Changing Models is optional when the current Model meets your needs.
Constantly changing Models can make it harder to understand why outputs differ.
A better comparison method is:
Same Tool
↓
Same Prompt
↓
Model A
↓
Review Result
↓
Model B
↓
Review Result
This lets you compare Models under similar conditions.
Two Models can respond differently to the same request.
Differences can include:
Therefore, selecting another Model can be useful when the current result does not meet the objective.
Before changing Models, read the description displayed beneath each Model Name.
For example, source-defined descriptions distinguish characteristics such as:
Use those descriptions as a first-level guide.
A provider can offer one or more Models.
For example:
Provider
Model
→ A particular Gemini model
or:
Provider
→ Anthropic
Model
→ A particular Claude model
Your API Key is generally associated with the provider rather than with one individual PROS SUITE Tool.
The current source-defined catalogue includes several Models routed through the NVIDIA provider family, including models originating from different AI organizations.
For users, this means the Model's branding and the provider used to access it are not always identical.
For example, a Model may have a familiar model-family Name while its provider in PROS SUITE is NVIDIA.
Follow the provider shown/required by PROS SUITE when configuring access.
If PROS SUITE reports:
Model Not Available
use the provider relationship presented by the Suite.
Do not assume a Key based solely on the Model's brand Name.
The current Model catalogue explicitly records both:
Settings includes a:
Current Status:
area for configured provider Keys.
It can display statuses such as:
🟢 Custom Key Set
or:
⚪ Not Configured
For Gemini, the current interface can also display:
🟠 using default/demo.
This can help you understand which personal provider credentials you have already configured.
There is no requirement for a normal user to enter:
all at once.
Configure only those required for Models you actually intend to use.
A good first-time workflow is:
Use Current Available Model
↓
If it works:
Continue
If you deliberately select another Model and receive:
Model Not Available
↓
Configure the appropriate provider Key
This avoids unnecessary credential configuration.
When you configure provider credentials in Settings, select:
Save Changes.
PROS SUITE confirms:
API Keys configuration saved
and:
“Your API keys have been updated successfully.”
Then return to Model selection.
If you opened Settings but decide not to configure another provider, select:
Cancel.
You can continue using an already available Model.
The workflow is:
Configure Provider Key
↓
Save Changes
↓
Model List Refreshes
↓
Open Select AI Model
↓
Select Desired Model
Entering an API Key does not itself necessarily mean that Model has become your currently selected Model.
When you manually select a Model, PROS SUITE records the choice as a manual selection and stores it locally in the browser.
The browser can then restore that selection when PROS SUITE loads again.
Because the available Model collection can change, a Model previously saved in your browser may later:
If your Header no longer shows the expected Model correctly:
Plan access determines whether you can use a PROS SUITE Tool.
Model availability determines which AI Model can be selected for supported Tool operations.
For example:
Business Plan included in your Plan
does not automatically mean:
Every AI Model is available without provider credentials.
These are separate access layers.
Having a large Token Balance does not automatically satisfy a missing provider API Key requirement.
You can have:
10,000 Tokens
and still receive:
Model Not Available
if the selected Model requires provider access that is not currently configured.
Likewise, having a provider API Key does not replace the need for sufficient PROS SUITE Tokens.
Before an AI operation, think of three layers:
Can your PROS SUITE account use the Tool?
Is the selected Model currently available/configured?
Do you have sufficient PROS SUITE Tokens for the operation?
The ideal state is:
Tool Available + Model Available + Sufficient Tokens
→ Generate.
Suppose Content Writer is already working normally.
Recommended workflow:
You do not need to change Models simply because the Selector exists.
Suppose an OpenAI-provider Model appears in the dropdown.
If an Anthropic Model is available:
The Settings interface identifies Anthropic credentials for Claude-family Models.
If you select a Model whose provider is NVIDIA:
The current Settings interface contains a dedicated NVIDIA provider field.
Suppose you want to compare writing quality.
Tool:
Content Writer
Prompt:
Use one complete fixed Prompt.
Model:
Model A
Save the result.
Keep the same Tool and Prompt.
Change to:
Model B
Recheck Token cost.
Generate again.
Then compare:
This creates a meaningful comparison.
If you do not see the Selector:
The current Header only renders the Model control when AI-settings visibility is enabled.
The ordinary subscriber does not need to manage the platform-level visibility control.
If Loading models... remains unusually persistent:
If you see:
Model Not Available
“This model requires an API key. Please configure it in Settings.”
the normal resolution is:
Settings → API Keys → Correct Provider → Enter Key → Save Changes → Select Model Again.
Try:
The available Model list is refreshed when API Key state changes.
Check the Model's provider.
For example:
Anthropic Model
needs Anthropic provider access—not OpenAI.
Return to:
Settings → API Keys
and configure the correct provider.
This can happen because the manually selected Model is stored in the browser.
Open Select AI Model and choose your preferred Model again.
A previously available Model may no longer appear because the list is dynamically loaded.
Do not attempt to force an old Model Name manually.
Instead:
If you are using a standard ToolPage and changed to a recognized OpenAI-provider Model, the current Token display can increase according to its verified OpenAI multiplier.
This is expected behavior in the shared standard generator.
Recheck the cost before generating.
This can occur because PROS SUITE contains different AI architectures.
A standard generator may use the selected Model directly through the shared streaming workflow, while a specialized Tool may use additional or separate generation processes.
Follow the dedicated chapter for that Tool.
Provider access does not replace PROS SUITE Tokens.
If the Model is available but the operation reports insufficient Tokens:
This is a Token issue, not a Model-credential issue.
Likewise, sufficient Tokens do not replace provider access.
If you have enough Tokens but see:
Model Not Available
configure the required provider or use another Model already available.
New users should generally begin with the Model already selected by PROS SUITE.
Learn:
before adding multiple provider credentials and comparing many Models.
Good reasons include:
Avoid switching merely because many Model Names are available.
When using a standard generator:
Model Change
↓
Provider May Change
↓
Displayed Cost May Change
↓
Review Cost Before Generate
This prevents unexpected Token usage.
Adding every available provider Key gives little benefit if you do not intend to select those Models.
Configure provider access as needed.
This keeps your Setup simpler and makes Model selection easier to understand.
If testing quality, keep constant:
Then change only the Model.
Otherwise, you cannot determine whether the difference resulted from the Model or your changed instructions.
Before generating the same request using another Model:
This prevents accidentally losing a result you preferred.
Descriptions help summarize a Model's intended strengths, but actual performance depends on:
Test Models on your actual workflow before deciding which one you prefer.
Your API Keys are sensitive credentials.
Do not:
Enter them only through the appropriate Settings → API Keys fields when needed.
The complete user workflow is:
Open PROS SUITE
↓
Header
↓
Select AI Model
↓
Review Current Available Models
↓
Read Name + Description
↓
Choose Model
↓
If available:
Selected
If unavailable:
Model Not Available
↓
Settings → API Keys
↓
Configure Correct Provider
↓
Save Changes
↓
Model List Refreshes
↓
Select Model Again
↓
Return to Tool
↓
Check Token Cost
↓
Generate
Before changing Models, ask:
If a Model cannot be used:
Ordinary PROS SUITE users do not need this manual to explain how to:
Those are not subscriber operations.
The user needs to know only:
Which Models are available to me?
Which one am I using?
Can I select another one?
Does it require my provider Key?
Does switching affect my Token cost?
Several source-verified details are worth keeping together:
The current source default is:
Gemini 3.6 Flash.
The actual available Model collection can be refreshed dynamically.
A manually selected Model is saved in the browser.
If gemini-2.5-flash appears in the Model list, the Header currently permits direct selection without applying its normal provider-key availability check.
The selector checks whether the required provider access is available and otherwise opens Settings with Model Not Available.
For these reasons, the live Selector should always be considered the best source for current user Model availability.
The user-facing model system can be understood as:
TOOL
What do I want to accomplish?
↓
MODEL
Which available AI engine should perform the supported AI work?
↓
PROVIDER ACCESS
Is the required Model/provider available?
↓
TOKEN COST
What does the Tool/action currently cost?
↓
PROMPT
What do I want generated?
↓
OUTPUT
Review the result.
The three important access layers remain:
Tool Access + Model Access + Token Availability
AIUNIFY PROS SUITE provides a shared AI Model Selector in the Header when that functionality is available.
The Selector displays the currently selected Model, opens under:
Select AI Model
and presents dynamically available Models with their Names and descriptions.
The current source recognizes provider families including:
Google, OpenAI, Anthropic, Mistral, Groq, NVIDIA, and Other.
The source-defined default Model is currently Gemini 3.6 Flash, but the available Model collection can be refreshed dynamically, so users should rely on the live Model Selector rather than a permanent list in this manual.
When a Model requires provider access that is not available, PROS SUITE displays:
Model Not Available
“This model requires an API key. Please configure it in Settings.”
and opens Settings.
A manually selected Model is remembered in the browser.
For standard ToolPage generators, Model choice can also affect the displayed Token cost. A recognized OpenAI-provider Model currently applies a 1.5× displayed-cost multiplier, rounded upward, to the standard Tool's base cost.
The recommended operating sequence is:
Choose the Right Tool → Use the Current Model Unless You Have a Reason to Change → Select Another Model When Appropriate → Configure the Correct Provider Only if Required → Recheck Token Cost → Generate → Compare Results