AIUNIFY PROS SUITE can work with AI Models from multiple provider families. Some Models may already be available through the PROS SUITE environment, while others can require you to provide your own valid API Key for the applicable AI provider.
The subscriber-facing API Keys section currently supports:
This chapter covers:
This chapter covers only API Keys entered and managed by the individual PROS SUITE subscriber. It does not document system-wide or administrative AI provider configuration.
An API Key is a credential issued by an AI provider that can authorize use of that provider's services.
Examples of provider families currently supported by PROS SUITE include:
PROS SUITE uses provider-specific credentials rather than one universal Key that works with every Model.
A normal subscriber does not need to configure all six providers.
For example, if your current PROS SUITE Model works properly and you have no reason to select another provider:
No additional API Key configuration may be necessary.
Configure a provider Key when:
Your:
PROS SUITE Password
and:
AI Provider API Keys
are completely different credentials.
Your PROS SUITE Password authenticates you to your AIUNIFY account.
An AI provider API Key authorizes access to an external AI provider for supported AI operations.
Never enter your PROS SUITE Password into an API Key field.
Provider API access does not replace your PROS SUITE Token Balance.
You can have:
Valid Provider API Key + Insufficient PROS SUITE Tokens
and still be unable to complete a Token-consuming operation.
Likewise:
Sufficient PROS SUITE Tokens + Missing Provider API Key
can result in:
Model Not Available
These are separate requirements.
When using a provider-specific Model, think of the workflow in three layers:
Is the PROS SUITE Tool available to your account?
Can the selected AI Model be used?
Do you have enough PROS SUITE Tokens for the operation?
The desired condition is:
Tool Available + Model Available + Sufficient Tokens
↓
Generate
Open:
User Menu → Settings
The Settings dialog is titled:
Settings
with:
“Manage your API keys and security settings.”
When API Key controls are available, Settings contains:
Select:
API Keys
On supported desktop layouts, the Header also includes a Settings icon that opens the same Settings dialog.
You can therefore use:
User Menu → Settings
or, when visible:
Header → Settings icon
You may not need to open Settings manually.
If you select an AI Model that requires credentials that are not currently available, PROS SUITE can automatically open Settings and display:
Model Not Available
with:
“This model requires an API key. Please configure it in Settings.”
This gives you a direct path from Model selection to provider configuration.
The subscriber API Keys tab only appears when AI Settings are currently exposed in the PROS SUITE user interface.
If Settings contains Security (2FA) but no API Keys tab:
When API Keys is available, it contains six credential fields:
It also provides:
Each provider credential field currently uses a Password-style input.
For example:
Gemini API Key
uses:
type="password"
The same pattern is used for the other provider Key fields.
Therefore, entered credentials are concealed visually rather than displayed as ordinary plain text.
Unlike the Login Password field, the current API Key Settings source does not establish a separate:
Show / Hide
button for provider credentials.
The Key fields themselves remain Password-type inputs.
This manual therefore does not instruct users to reveal API Keys from this screen.
When you open the Settings dialog, PROS SUITE loads the provider credentials currently stored in your browser into the corresponding fields.
Because the fields are Password-type inputs, the values remain visually masked.
The first provider field is:
Gemini API Key
with:
Enter Gemini API key
The current user interface describes it as being required for Gemini and other Google Models.
Use this field when the Google/Gemini Model you intend to use requires your own provider credential.
The next field is:
OpenAI API Key
with:
Enter OpenAI API key
The interface identifies it for OpenAI Models.
Use this field when the selected OpenAI Model requires your own OpenAI credential.
The next field is:
Anthropic API Key
with:
Enter Anthropic API key
The current interface associates this credential with Claude-family Models.
The next field is:
Mistral API Key
with:
Enter Mistral API key
The current interface associates this provider credential with Mistral/Mixtral functionality.
The next field is:
Groq API Key
with:
Enter Groq API key
The current interface specifically identifies Groq access for Llama Models routed through Groq.
The final current field is:
NVIDIA API Key
with:
Enter NVIDIA API key
The interface identifies it for Models accessed through the NVIDIA provider.
As explained in Chapter 10, the Model's brand Name and the provider through which PROS SUITE accesses it are not always the same.
Do not determine the required API Key solely from the Model's Name.
For example, a Llama-family or another third-party Model can be routed through:
The current Model definitions record a separate provider for each Model.
When PROS SUITE requires credentials, use the provider corresponding to the Model currently presented through the Suite.
PROS SUITE does not issue the external provider credential through this Settings form.
You need a valid API Key from the applicable provider according to that provider's own account and credential process.
Once you possess the Key:
Provider → Valid API Key → PROS SUITE Settings → Correct Provider Field
Do not enter a Key belonging to a different provider.
To add a provider credential:
After entering or changing provider credentials, select:
Save Changes.
PROS SUITE saves all six current field values into its user-side provider credential state.
After selecting Save Changes, PROS SUITE displays:
API Keys configuration saved
with:
“Your API keys have been updated successfully.”
The Settings dialog then closes.
If you open API Key Settings but do not want to save the current changes, select:
Cancel.
The dialog closes without running the Save action.
Use Cancel if:
Below the provider fields, PROS SUITE provides:
Current Status:
This section gives you a quick indication of whether a user-entered credential currently exists for each provider.
When a provider field contains a user-entered credential, its status appears as:
🟢 Custom Key Set.
For example:
OpenAI: 🟢 Custom Key Set
indicates that an OpenAI Key is currently present in the user-side Settings state.
For OpenAI, Anthropic, Mistral, Groq, and NVIDIA, an empty user credential displays:
⚪ Not Configured.
This means no user-entered Key is currently saved in that provider field.
It does not necessarily prove that no Model from that provider can ever be available, because PROS SUITE can also recognize provider access already available through its environment.
Gemini behaves slightly differently in the Current Status panel.
When you have entered your own Gemini Key:
Gemini: 🟢 Custom Key Set
When the user field is empty:
Gemini: 🟠 using default/demo.
Therefore, an empty Gemini user field does not display Not Configured in the current interface.
The status:
🟠 using default/demo
means the user has not currently entered a custom Gemini credential into this Settings field.
It should not be interpreted as a permanent guarantee that every Google/Gemini Model will always be available without further configuration.
Model availability remains determined by the current PROS SUITE Model system.
The Current Status area updates according to whether the corresponding current input field contains a value.
This means the status is useful for answering:
“Have I entered a custom Key for this provider?”
It is not a complete health test of the provider credential.
The Current Status indicator is based on whether a value exists in the field.
Therefore:
🟢 Custom Key Set
means a credential has been entered.
It does not, by itself, verify:
Actual use of the selected Model is the practical test.
To replace a provider Key:
The Save action replaces the current provider value with the content currently in that field.
The current source does not provide a separate:
Delete Key
button.
Instead, you can remove the stored user value by:
The Save operation writes the current field values, including an empty value, back to the stored provider state.
After clearing and saving:
The status returns to:
⚪ Not Configured
The status returns to:
🟠 using default/demo
If a Model depended on your user-provided provider Key, removing that Key can make that Model unavailable.
The next time you attempt to select it, PROS SUITE may display:
Model Not Available
and ask you to configure the provider again.
Do not remove an API Key that you still need for active Model workflows.
The current client source stores the user-supplied provider Key collection in browser local storage.
When a Key is changed, the provider collection is updated and saved again locally.
For subscribers, the key operational consequence is:
Your custom API Key configuration is browser-specific.
Because these credentials are stored in the browser, configuring them on one device does not establish that they will automatically appear on another.
For example:
Office Computer → OpenAI Key configured
does not necessarily mean:
Laptop → OpenAI Key configured
On another browser/device, you may need to enter the credential again.
If you clear the browser's site/local storage, locally saved provider credentials can be removed.
Possible effects include:
Keep your original provider credential securely accessible through the provider's approved credential-management process.
Because the configuration depends on browser-local storage, temporary/private browsing sessions should not be relied upon as permanent API Key storage.
If the browser removes that session's local data when it closes, you may need to configure the credential again.
The Settings interface currently displays the following note:
“Keys are stored locally in your browser and are never sent to our servers.”
The first part—local browser storage—is confirmed by the client source.
However, there is an important implementation detail users should understand.
The same included client source shows non-empty provider Keys being placed into request headers when PROS SUITE requests the current Model list.
Supported AI-generation requests likewise include the available provider credentials in request headers sent to the PROS SUITE AI endpoint.
Therefore, the literal phrase “never sent to our servers” should not be interpreted to mean the Key never transits through PROS SUITE request endpoints.
The source bundle used for this User Operations Manual intentionally excludes backend API implementation.
Because of that, the supplied source confirms:
The supplied source does not establish:
For that reason, this manual will not make unsupported claims beyond what the current source verifies.
Regardless of storage implementation, API Keys should be treated as sensitive.
A provider Key can potentially be associated with:
Protect it accordingly.
Do not enter a provider credential into:
Use only:
Settings → API Keys
for subscriber provider credential configuration.
Because the Settings fields are masked, avoid taking actions that expose the actual Key before capturing a screenshot.
Do not include Keys in:
unless your organization's approved secure support process specifically requires appropriate protected handling.
Each user should use credentials they are authorized to use.
Do not casually provide your personal or business provider API Key to another subscriber merely because they cannot access a Model.
A provider credential can have consequences outside PROS SUITE, including usage associated with the underlying provider account.
Your API Key security is only part of the credential chain.
Use appropriate security on the provider account that issues the Key, such as:
Follow the provider's own credential-management practices.
PROS SUITE does not itself provide a provider-Key regeneration function.
If a provider API Key is exposed or compromised, the Key must be managed through the provider that issued it.
After obtaining a replacement credential:
When the provider Key collection changes, PROS SUITE refreshes its Model list.
This allows newly configured provider access to affect the Models presented by:
Select AI Model
The typical process is:
Enter Provider Key
↓
Save Changes
↓
PROS SUITE Updates User Credential State
↓
Model List Refreshes
↓
Open Select AI Model
↓
Select Desired Model
When refreshing the Model collection, PROS SUITE adds each non-empty provider credential to a provider-specific request header.
This allows Model availability to be evaluated according to the provider credentials currently configured by the user.
The shared AI-generation workflow also takes the currently available API Keys and includes them with the AI request when present.
This is how supported workflows can use the provider credentials you have configured.
Provider Keys are stored by provider rather than by individual Tool.
For example, a configured OpenAI Key can be available to supported PROS SUITE workflows using an OpenAI provider Model.
You do not need a separate OpenAI Key for:
solely because they are different PROS SUITE tools.
The Model/provider relationship governs credential use.
Although the provider Key is shared at the provider level, this does not guarantee that every PROS SUITE Tool supports every Model identically.
Specialized applications can use:
Follow each Tool's dedicated chapter when necessary.
After configuring a provider Key:
Key Configured ≠ Model Selected
You still need to:
A provider API Key does not bypass your PROS SUITE subscription.
For example:
Research Agent not included in your account
Valid OpenAI API Key
does not automatically unlock Research Agent.
Tool access remains controlled separately from provider credentials.
Likewise:
Save OpenAI Key
does not increase:
Monthly Tokens
Provider credentials and PROS SUITE Token allocations are independent.
After configuring a provider and selecting its Model, remember Chapter 9's Token behavior.
For the standard ToolPage, an OpenAI-provider Model currently causes the displayed base cost to be multiplied by 1.5 and rounded upward.
Therefore:
Configure Key → Select Model → Recheck Token Cost → Generate
Suppose you select an OpenAI Model and receive:
Model Not Available
Settings opens.
Select API Keys.
Locate:
OpenAI API Key
Enter your valid credential.
Confirm Current Status changes to:
🟢 Custom Key Set
Select:
Save Changes
Return to:
Select AI Model
Select the desired OpenAI Model again.
Return to your Tool.
Recheck:
tokens per use
before generating.
To configure an Anthropic Model:
Settings
↓
API Keys
↓
Anthropic API Key
↓
Enter Key
↓
Custom Key Set
↓
Save Changes
↓
Select AI Model
↓
Select Anthropic Model
↓
Generate
The current interface associates the Anthropic field with Claude-family Models.
For a Model accessed through NVIDIA:
Settings → API Keys → NVIDIA API Key
Enter the authorized provider credential and select:
Save Changes.
Then return to the Model Selector and select the desired NVIDIA-provider Model.
If you no longer want your user-entered OpenAI Key stored in this browser:
Models that depended on that credential may no longer be selectable.
If a Model previously worked but the provider credential has been replaced or revoked:
If you see:
Model Not Available
“This model requires an API key. Please configure it in Settings.”
check:
Entering a credential into the field alone is not enough to persist the change.
Select:
Save Changes
If you close with Cancel, the Save handler is not executed.
If a provider still shows:
⚪ Not Configured
This is the current normal empty-user-key status for Gemini:
🟠 using default/demo.
If your desired Google/Gemini Model already works, you may not need to change anything.
If the desired Model specifically reports that an API Key is required, enter an appropriate Gemini credential and save it.
Remember that:
Custom Key Set
only confirms that a value exists.
Possible causes of continued failure include:
Read the actual Model or generation error before changing other settings.
Because API Key settings are stored in browser-local storage, the other computer's configuration does not automatically transfer.
Configure the authorized provider Key in the current browser as needed.
If previously configured Keys are no longer present, consider whether:
Re-enter the necessary provider credentials through Settings.
Suppose you selected a Model whose provider is NVIDIA but entered an OpenAI Key.
The Model can remain unavailable because the credential does not match the provider relationship.
Return to:
Settings → API Keys
and configure the correct provider.
If Settings does not show:
API Keys
the user-facing API Key controls are not currently exposed in your PROS SUITE interface.
Do not attempt to access administrative settings.
Use the Models currently available or contact the appropriate AIUNIFY support channel if a required subscriber capability appears unavailable.
If the provider Model is available but generation reports insufficient Token Balance:
This is not an API Key problem.
Check:
Monthly Tokens
and the operation's Token cost.
Provider authentication and PROS SUITE Token usage remain separate.
If you believe a provider API Key has been exposed:
Do not attempt to “change” an external provider Key solely by editing its text in PROS SUITE; the actual credential must be issued by the provider.
A clean subscriber setup may contain:
Gemini → default/demo
OpenAI → Custom Key Set
Anthropic → Not Configured
Mistral → Not Configured
Groq → Not Configured
NVIDIA → Not Configured
if OpenAI is the only additional provider you actually use.
There is no benefit to filling every field simply because it exists.
Choose a Model first based on your workflow.
Then configure its provider only if necessary.
Recommended sequence:
Business Task
↓
Correct PROS SUITE Tool
↓
Desired Model
↓
Provider Required?
↓
If yes:
Configure that Provider Only
↓
Generate
Before using:
review your content to ensure you have not accidentally included an API Key in the Prompt or generated output.
Credentials should never become part of client-facing content.
Do not depend on PROS SUITE browser storage as your only record of an important provider credential.
Maintain authorized credential management through the provider and your organization's approved secure processes.
This is particularly important because browser data can be deleted.
Because credentials are browser-local, if you stop using a device for PROS SUITE:
This is especially important for shared or reassigned devices.
Entering a business or personal provider API Key into a shared browser can expose that credential to other people who have access to the same browser profile or device.
Prefer a private, controlled device and browser profile.
PROS SUITE displays its own:
Monthly Tokens
But your external provider account may maintain separate usage, billing, quotas, or limits.
Those provider-side metrics are not the same as PROS SUITE Tokens.
When using your own provider credentials, maintain awareness of both systems where applicable.
The complete subscriber workflow is:
Select AI Model
↓
If available:
Use Model
If not:
Model Not Available
↓
Settings
↓
API Keys
↓
Identify Provider
↓
Enter Valid API Key
↓
Current Status → Custom Key Set
↓
Save Changes
↓
Model List Refreshes
↓
Select AI Model
↓
Choose Model Again
↓
Check Tool Token Cost
↓
Generate
Settings
↓
API Keys
↓
Locate Existing Provider
↓
Replace Current Masked Value
↓
Save Changes
↓
API Keys configuration saved
↓
Model List Refresh
↓
Retry Model
Settings
↓
API Keys
↓
Clear Provider Field
↓
Save Changes
↓
For most providers:
Not Configured
or for Gemini:
using default/demo
↓
Previously Key-Dependent Models May Become Unavailable
Before saving a provider credential:
You should now be able to:
This User Operations Manual does not instruct subscribers how to:
Those are not ordinary PROS SUITE user operations.
The subscriber needs only to understand:
Do I need my own provider Key?
Which provider does this Model use?
Where do I enter it?
How do I save or remove it?
Is the Model now available?
The current Settings interface tells users:
“Keys are stored locally in your browser and are never sent to our servers.”
The source does verify that subscriber Keys are stored in browser-local storage.
The same source also verifies that non-empty provider credentials are included as headers in requests to the application's Model and AI-generation endpoints.
Because the backend API implementation was intentionally excluded from the sanitized User Operations source bundle, this manual cannot establish what those endpoints do with the credential after receiving the request.
Accordingly, users should treat all configured provider API Keys as sensitive credentials transmitted as necessary for the applicable AI workflow, rather than interpreting the interface wording as a guarantee that a Key never leaves the browser during use.
The subscriber model is:
AIUNIFY PROS SUITE ACCOUNT
↓
PROS SUITE TOOL
↓
SELECTED AI MODEL
↓
MODEL PROVIDER
↓
If provider already available:
USE MODEL
If personal credential required:
SETTINGS → API KEYS
↓
ENTER PROVIDER KEY
↓
SAVE CHANGES
↓
CUSTOM KEY SET
↓
SELECT MODEL
↓
CHECK PROS SUITE TOKENS
↓
GENERATE
Remember:
API Key = Provider Access
PROS SUITE Tokens = AIUNIFY Usage Balance
Plan = Tool Access
These three concepts should remain distinct.
AIUNIFY PROS SUITE provides a subscriber-facing API Keys workspace through:
Settings → API Keys
The current interface supports user credentials for:
Gemini, OpenAI, Anthropic, Mistral, Groq, and NVIDIA.
Each field is a masked Password-type input. Existing locally stored user credentials are loaded into Settings when the dialog opens.
The Current Status section identifies:
🟢 Custom Key Set
when a user credential is present, and:
⚪ Not Configured
for empty OpenAI, Anthropic, Mistral, Groq, and NVIDIA fields. Gemini instead displays:
🟠 using default/demo
when no custom Gemini Key has been entered.
Selecting Save Changes updates the provider credential collection and displays:
API Keys configuration saved
“Your API keys have been updated successfully.”
Manually entered provider Keys are stored in browser-local storage, meaning they are browser/device-specific.
When provider credentials change, PROS SUITE refreshes the available Model collection.
When a required provider Key is missing, Model selection can display:
Model Not Available
“This model requires an API key. Please configure it in Settings.”
The recommended operating sequence is:
Choose Tool → Choose Model → Configure Provider Only if Required → Save Changes → Select Model Again → Recheck Token Cost → Generate
Provider API Keys, PROS SUITE Tokens, and Plan Tool Access are three separate parts of the overall AIUNIFY PROS SUITE operating model.