Documentation/Getting Started

Getting Started

Set up your KeyRing AI account, activate your license, connect your AI providers, and send your first multi-model prompt.

Read time: 6 min read

Platform Availability

KeyRing AI is currently available for Windows 10+ (64-bit). macOS and Linux builds are in development.

What is KeyRing AI?

KeyRing AI is a Windows desktop application that lets you run multiple AI providers from one interface. Compare answers side by side, run multi-AI discussions, attach files for context, and track your usage — all without switching between different provider websites.

What You Can Do

Query up to 11 AI providers at once (plan dependent)
Compare model output in Chatroom or individual provider tabs
Run Roundtable sessions where multiple AIs discuss the same topic
Build and run custom AI agents (Pro tier)
Your API keys stay on your machine — we never see them

Your Getting Started Journey

Here is the full path from sign-up to your first prompt, broken into three parts:

1. Get Your License

2. Install & Activate

Download the app, paste your key, and you are in.

3. Connect & Go

Add your provider API keys and send your first prompt.

Part 1: Get Your License Key

Create your account

Head to pricing and click Join Beta Program. You will need an invite code to register — if you do not have one yet, you can join the beta waitlist.

Verify your email

After signing up, check your inbox for the verification email and click the link to confirm your account.

Reserve your beta seat

Once verified, go to the beta page, choose your tier, and complete checkout.

Copy your license key from the dashboard

After payment, your license key is ready automatically. Go to your dashboard, find your key, and copy it. You will need it in the next step.

Pro Tip

You will also receive a welcome email confirming your purchase. Your license key is always available in your dashboard.

Part 2: Install and Activate

Download the installer

Go to the download page and grab the Windows installer.

Install the app

Run the installer and follow the setup steps. When it finishes, launch KeyRing AI.

First Launch Takes a Moment

The first time you open KeyRing AI, it takes a bit longer while everything initializes behind the scenes. This is normal and only happens once.

Paste your license key

When the app opens, you will see the onboarding screen. Paste the license key you copied from your dashboard and click Activate License.

Add your first provider API key

After your license activates, the next screen asks you to add an API key for an AI provider. Pick a provider, paste its key, and click Save & Verify. Or skip this step and do it later in API Settings.

Part 3: Connect Providers and Send Your First Prompt

Need a beginner-friendly walkthrough?

Check out API Keys Made Easy for a plain-language, copy-and-paste walkthrough.

Open API Settings

Click the API Settings button in the left sidebar to open the full configuration panel.

Add your provider API keys

Find the provider you want, paste in your API key, and click Save Changes. Repeat for any other providers you want to use.

Your Keys Stay on Your Machine

Provider API keys are encrypted and stored locally on your computer. KeyRing uses them directly from your desktop to call each provider's API — your keys never pass through our servers.

Provider Terms Apply

By providing API keys, you connect directly to OpenAI, Google, Anthropic, xAI, Mistral, Groq, Cohere, DeepSeek, Together AI, Perplexity, ElevenLabs services. Usage is subject to each provider's terms of service and privacy policies. KeyRing AI does not provide these AI services directly and is not responsible for content generated by those models.

Choose your models

Each provider card has a model dropdown. Pick the model you want to use. The app sticks with your selection — it does not auto-switch.

Enable providers and send a prompt

In the main workspace, turn on the providers you want in Controls - Providers, type your prompt, and hit Submit (or press Ctrl+Enter).

text

Compare three approaches to reduce cold start latency in desktop AI apps. Include tradeoffs and a recommended implementation order.

Compare the results

Switch between the Chatroom tab and individual provider tabs to see how each model answered. If you enabled Consensus, check that tab too for a combined answer.

Attachments Quickstart

Click the paperclip icon in the left sidebar to open the Attachments module. From there you can add files and decide which providers should receive them.

Choose a scope: Global (all providers) or specific to one provider
Pick how much of each file to send: full text, first few lines, metadata only, or filename only
Set a character limit per scope to control how much context gets sent
Add files by clicking the picker or dragging and dropping

Troubleshooting

Common Issues

App feels slow on first launch: This is normal. The backend initializes on first run and takes a bit longer.
Provider not responding: Double-check your API key is correct, confirm you have active billing with that provider, and verify the selected model is accessible on your account.
No output in a tab: Make sure that provider is turned on in Controls and that your plan tier includes it.

Still stuck? Contact support.