Feature Guide Overview
Glacier AI puts models from 40+ providers — OpenAI, Claude, Gemini and others — behind a single OpenAI-compatible API. With one account and one API key you can call every model from the same address, and see all of your usage and billing in one place.
This guide covers everything you will use day to day: registering, creating a key, calling models, reading logs, and managing your balance and subscriptions.
Where to start
Section titled “Where to start”On your first visit, read these in order — each step depends on the one before it:
- Sign Up & Sign In — get an account.
- API Keys — get a calling credential.
- Using the API & Playground — make your first call.
- Wallet & Subscriptions — make sure you have quota.
- Logs & Async Tasks — confirm usage and charges.
If you only want a working call as fast as possible, go straight to the Quick Start.
All features
Section titled “All features”Registration, username and password sign-in, password reset, plus one-click sign-in with passkeys and third-party OAuth.
- Username and password sign-in
- Passkey and third-party OAuth sign-in
- Registration flow
- Login session management
- Forgotten password reset
The overview page summarizes balance, usage, success rate and latency. The dashboard breaks calls down by time, model and group, and shows how requests are distributed.
- Quick start wizard
- Usage and health status
- API information, announcements, FAQ and uptime
- Model call analytics
- Request flow
Manage account information, email and OAuth bindings, notification channels, interface language and daily check-in, plus security settings: password, two-factor authentication (2FA), passkeys and login sessions.
- View account information
- Manage account bindings
- Set notification and behavior preferences
- Set the interface language preference
- Personalize the sidebar
- Daily check-in
- Security settings
- Enable or disable two-factor authentication (2FA)
- Enable or disable passwordless passkey sign-in
- Login session management
An API key is the credential used to call the API. Each key can have its own quota, expiry, model restrictions, group and IP allowlist. Create one per use case rather than sharing a single key everywhere.
- Create an API key
- View and copy an API key
- Edit an API key
- Enable, disable and delete API keys
Try models directly in the playground, or use an API key to call chat, Claude, Gemini, image and audio endpoints through the OpenAI-compatible API.
- Playground
- API address and authentication
- SDK samples
- Supported endpoints
Import your configuration into popular AI chat applications and use the platform’s models without writing any code.
- One-click import from a chat preset
- One-click import from an API key
- Manual chat app configuration
Inspect calls one by one — tokens, latency and billing breakdown. Video, music and drawing tasks run asynchronously and track their own status, results and refunds on failure.
- Usage logs
- Drawing logs
- Async task logs
- Task status and background processing
- API reference: log query endpoints, async generation and task endpoints
Check your balance and usage, top up online, redeem codes, buy subscription plans, choose the deduction order, claim referral rewards and review order history.
- View balance and usage
- Online top-up
- Redemption codes
- Subscription plans
- Buying a subscription plan
- Checking subscription status
- Choosing a billing preference
- Referral rewards
- Order history
- Common limitations
See each model’s current price, capabilities and runtime status, and compare how heavily models and providers are used. Checking here before you call avoids most errors.
- Model marketplace
- Rankings