Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/TabbyAIKeyboard/tabby/llms.txt

Use this file to discover all available pages before exploring further.

Tabby supports multiple AI providers through the Vercel AI SDK, allowing you to choose the best models for your use case. Configure your preferred providers by adding API keys to your environment variables.

​
Supported Providers

Tabby integrates with the following AI providers:
  • OpenAI - GPT models for chat, coding, and memory operations
  • Groq - Fast inference with Llama and other open models
  • Cerebras - High-performance AI inference
  • Google Generative AI - Gemini models
  • OpenRouter - Access to multiple models through a single API
  • XAI - Additional AI capabilities

​
Configuration

​
Next.js Backend

The Next.js backend (nextjs-backend/.env.local) handles AI provider configuration for the application’s main features. 
# AI Providers
OPENAI_API_KEY="sk-..."
GOOGLE_GENERATIVE_AI_API_KEY="..."
GROQ_API_KEY="gsk_..."
CEREBRAS_API_KEY="..."
OPENROUTER_API_KEY="..."

# Tools
TAVILY_API_KEY="tvly-..."

​
Memory Backend

The Python memory backend (backend/.env) requires an OpenAI API key for memory operations with Mem0. 
OPENAI_API_KEY="sk-..."
The memory backend uses OpenAI’s gpt-4.1-nano-2025-04-14 model for memory classification and vision capabilities.

​
Provider Setup

​
OpenAI

  1. Visit OpenAI Platform
  2. Create an account or sign in
  3. Navigate to API Keys section
  4. Click Create new secret key
  5. Copy the key (starts with sk-)
OpenAI API key is required for the memory backend. The application will not function properly without it.

​
Groq

  1. Visit Groq Console
  2. Create an account or sign in
  3. Navigate to API Keys
  4. Click Create API Key
  5. Copy the key (starts with gsk_)

​
Cerebras

  1. Visit Cerebras Cloud
  2. Create an account or sign in
  3. Navigate to API Keys
  4. Generate a new API key
  5. Copy the key

​
Google Generative AI

  1. Visit Google AI Studio
  2. Sign in with your Google account
  3. Click Create API Key
  4. Copy the generated key

​
OpenRouter

  1. Visit OpenRouter
  2. Create an account or sign in
  3. Navigate to Keys section
  4. Create a new API key
  5. Copy the key

​
AI SDK Integration

Tabby uses the Vercel AI SDK to integrate with providers. The SDK packages are already included in the project:
package.json
{
  "dependencies": {
    "@ai-sdk/openai": "^3.0.7",
    "@ai-sdk/groq": "^3.0.4",
    "@ai-sdk/cerebras": "^2.0.5",
    "@ai-sdk/google": "^3.0.6",
    "@ai-sdk/openai-compatible": "^2.0.10",
    "ai": "^6.0.23"
  }
}

​
Web Search with Tavily

Tabby supports web search capabilities through Tavily, enhancing AI responses with real-time information.
  1. Visit Tavily
  2. Create an account
  3. Navigate to the dashboard
  4. Copy your API key

​
Best Practices

​
API Key Security

  • Never commit .env files to version control
  • Use .env.example as templates for required keys
  • Rotate API keys regularly
  • Set up usage limits and alerts in provider dashboards

​
Provider Selection

  • OpenAI - Best for general-purpose tasks and memory operations (required)
  • Groq - Fastest inference, ideal for real-time interactions
  • Cerebras - High-performance for intensive workloads
  • Google Gemini - Alternative to OpenAI for vision and reasoning
  • OpenRouter - Access multiple models with a single API key

​
Cost Optimization

  • Monitor usage in provider dashboards
  • Use cheaper models for simple tasks (e.g., text formatting)
  • Reserve powerful models for complex tasks (e.g., coding assistance)
  • Consider rate limits and quotas for each provider

​
Troubleshooting

  • Verify the key is correctly copied (no extra spaces)
  • Check if the key has been activated in the provider dashboard
  • Ensure you have credits/billing set up
  • Restart the backend after adding new keys
  • Verify the API key is set in the correct .env.local file
  • Restart both the Next.js backend and Electron app
  • Check console logs for configuration errors
The memory backend requires an OpenAI API key. Ensure OPENAI_API_KEY is set in backend/.env.

​
Next Steps

Memory Backend

Configure persistent memory with Mem0

Settings

Customize application settings

Build docs developers (and LLMs) love

Get started for freeTalk to us