Frequently Asked Questions
Quick answers to the questions we hear most about Ask Sage — accounts, tokens, security, models, and troubleshooting.
Table of Contents
Getting Started
How do I sign up for Ask Sage?
Can I interface with Ask Sage through an API?
What URLs can I use to access Ask Sage?
- For FedRAMP High/IL5 DoD use (CUI authorized models only): https://chat.genai.army.mil (requires NIPR connection)
- For FedRAMP High/IL5 civilian use (CUI authorized models only): https://chat.civ.asksage.ai/
- For FedRAMP High and IL2 commercial use (non-CUI models): https://chat.asksage.ai/
- IL6 (Secret): Requires SIPR connection
- There are other URLs for specific customer deployments — contact support@asksage.ai for more information.
Subscriptions & Billing
Is Ask Sage free to use?
How do I manage my subscription?
Purchase/Modify: Click the "Manage Subscription" link in the Tokens tab of your Settings, or email billing@asksage.ai.
Upgrade: You can upgrade your plan anytime.
Cancel: Email billing@asksage.ai before your next billing cycle. No refunds are issued for cancellations mid-cycle.
How can I purchase more tokens?
Security & Data Privacy
How does Ask Sage protect my data from being used to train models?
Ask Sage utilizes "fire and forget" APIs to ensure your data is never used for training:
Infrastructure: We leverage Azure OpenAI APIs on Azure commercial FedRAMP high cloud (fully managed by Microsoft). Our vector database storing CUI is hosted in Ask Sage's Azure GovCloud at IL5.
Embeddings, Not Fine-Tuning: We embed information to maintain model agnosticism rather than fine-tuning models. This allows us to use your data with any model in the future.
Fire and Forget Queries: At query time, we only pass 3-6 relevant results to the Azure OpenAI model. Once a response is generated, the query data is not retained or stored.
This approach, combined with our secure infrastructure, guarantees that user data remains protected and is never used to train models.
Is my data accessible to other users?
What does "Gov" designation mean for models?
How do I limit my organization to CUI-compliant models only?
Datasets & Data Management
What is a dataset?
Can I share my datasets with other users?
What types of documents can I ingest?
Tokens & Usage
What are Ask Sage tokens?
Ask Sage tokens are model-agnostic and serve as the central currency for the platform. One token is roughly equivalent to 3.7 English characters.
Inference Tokens: Used to make predictions or generate responses. Consumed when you submit prompts, use plugins/agents, or interact with models.
Training Tokens: Used for creating and populating datasets. Consumed when uploading data as it's processed into embeddings and stored in the vector database.
For a deeper walkthrough, see Ask Sage Tokens.
When do tokens replenish and do they roll over?
Why do I see a different token count in the bottom right corner?
That badge opens the Conversation Context panel, which shows how much of the current model's context window your conversation is using. It is a per-conversation capacity set by the model — it resets to zero every time you start a new chat, and it varies by model.
It is not your Ask Sage token balance. To check your available tokens, go to the Tokens tab in your Settings.
For a full breakdown of the panel's fields, see Conversation Context.
Account & Login
Why am I getting a "Login Failed" message?
How do I add my CAC/PIV to my account?
Before signing in with your CAC/PIV, you must manually associate it with your account:
- Sign in with your Email and Password (not "Sign in with CAC/PIV")
- Click the Settings cog in the bottom left corner
- Select the Account tab
- Scroll to "Register CAC/PIV to your account" and click it (ensure your CAC is in the reader)
- Log out and sign back in using your CAC/PIV
How do I update my CAC/PIV?
How do I manage MFA (Multi-Factor Authentication)?
Enroll: Go to Account Settings after signing in with your email and password. We recommend using Microsoft Authenticator App.
Reset: Email support@asksage.ai to request an MFA reset. You'll receive confirmation and re-registration instructions.
Note for Admins: You cannot enforce MFA from the Admin Panel. Inform members to register MFA in their Account Settings and monitor MFA status in the Users tab of the Admin Panel.
Why is my account disabled?
Models & Features
What is a Persona?
How do I share a persona?
What is Temperature?
What models can generate images?
What does the Live/Live+ function do?
Can I train a custom model?
Why does the model forget things I said earlier in a long chat?
Every model can only hold a limited amount of content in view at once — its context window. Models are stateless between turns, so your whole conversation is re-sent with each new message, and as a chat grows it fills that window. Once usage approaches 100%, the oldest messages are dropped to make room for new ones. Nothing is deleted from your saved chat history, but the model can no longer see the dropped messages.
Click the token badge in the bottom-right corner of the chat window to open the Conversation Context panel and see how full the window is. Starting a new chat when the topic changes is the simplest fix — see Conversation Context.