Connect LLMs

Connect LLMs

Link your AI model API keys to give your workspace access to Claude, GPT, Gemini, or other providers.

Supported Models

Provider	Models	Status
Anthropic	Claude 3.5, Claude 3	Supported
OpenAI	GPT-4, GPT-4o, GPT-3.5	Supported
Google	Gemini Pro, Gemini Ultra	Supported
Local	Ollama, LM Studio	Via custom endpoint

Adding API Keys

1. Go to Build Mode → Settings → AI Models

2. Select provider

3. Enter your API key

4. Click Verify & Save

Keys are encrypted and stored securely. They're only decrypted within your workspace session.

Key Security

Aspect	How It Works
Storage	Encrypted at rest, tied to your wallet
Access	Only available inside your workspace
Visibility	Never sent to ZYBER servers in plain text
Deletion	Removed when you delete them or your account

Using Multiple Models

You can add keys for multiple providers and switch between them during a session. Useful for comparing outputs or using different models for different tasks.

Verification

After adding a key, ZYBER sends a minimal test request to verify it works. If verification fails:

Issue	Solution
Invalid key	Check for typos, regenerate if needed
Expired key	Generate a new key from provider
Rate limited	Wait and retry
Network error	Check workspace connectivity

Custom Endpoints

For self-hosted models or alternative providers, use Custom Endpoint:

1. Select "Custom" as provider

2. Enter the API endpoint URL

3. Add any required headers

4. Test connection