Supported Providers

Complete list of AI providers supported by Tokenlay, including configuration patterns and model availability.

Provider Overview

Tokenlay supports all major AI providers through a unified API interface. Each provider can be configured using specific key parameters or automatic inference.

OpenAI

Configuration

const client = new TokenlayOpenAI({
  openai_key: process.env.OPENAI_API_KEY,
  // or
  provider_api_key: "sk-...", // Auto-detected
  tokenlayKey: process.env.TOKENLAY_KEY,
});

Supported Models

• GPT-4o, GPT-4o-mini
• GPT-4, GPT-4 Turbo
• GPT-3.5 Turbo
• DALL-E 3, DALL-E 2
• Whisper, TTS

Key Pattern

• Prefix: sk-
• Example: sk-1234567890abcdef...

Anthropic

Configuration

const client = new TokenlayOpenAI({
  anthropic_key: process.env.ANTHROPIC_API_KEY,
  // or  
  provider_api_key: "sk-ant-...", // Auto-detected
  tokenlayKey: process.env.TOKENLAY_KEY,
});

Supported Models

• Claude 3.5 Sonnet
• Claude 3 Opus, Sonnet, Haiku
• Claude 2.1, Claude 2.0

Key Pattern

• Prefix: sk-ant-
• Example: sk-ant-1234567890abcdef...

Google

Configuration

const client = new TokenlayOpenAI({
  google_key: process.env.GOOGLE_API_KEY,
  // or
  provider_api_key: "AIza...", // Auto-detected
  tokenlayKey: process.env.TOKENLAY_KEY,
});

Supported Models

• Gemini 1.5 Pro, Flash
• Gemini 1.0 Pro
• PaLM 2

Key Pattern

• Prefix: AIza
• Example: AIzaSy1234567890abcdef...

AWS Bedrock

Configuration

const client = new TokenlayOpenAI({
  aws_access_key: process.env.AWS_ACCESS_KEY_ID,
  aws_secret_key: process.env.AWS_SECRET_ACCESS_KEY,
  aws_region: process.env.AWS_REGION,
  tokenlayKey: process.env.TOKENLAY_KEY,
});

Supported Models

• Claude 3.5 Sonnet (Bedrock)
• Claude 3 Opus, Sonnet, Haiku (Bedrock)
• Titan Text, Titan Embeddings
• Llama 2, Llama 3

Key Pattern

• Prefix: AKIA
• Example: AKIA1234567890ABCDEF

Azure OpenAI

Configuration

const client = new TokenlayOpenAI({
  azure_key: process.env.AZURE_OPENAI_KEY,
  azure_base_url: "https://your-resource.openai.azure.com",
  tokenlayKey: process.env.TOKENLAY_KEY,
});

Supported Models

• GPT-4o, GPT-4 (Azure deployments)
• GPT-3.5 Turbo (Azure deployments)
• DALL-E 3 (Azure deployments)

Key Pattern

• Custom format (varies by Azure setup)
• Requires explicit azure_key parameter

Cohere

Configuration

const client = new TokenlayOpenAI({
  cohere_key: process.env.COHERE_API_KEY,
  tokenlayKey: process.env.TOKENLAY_KEY,
});

Supported Models

• Command R+, Command R
• Command Light, Command
• Embed v3, Embed v2

OpenRouter

Configuration

const client = new TokenlayOpenAI({
  provider_api_key: process.env.OPENROUTER_API_KEY,
  provider_base_url: "https://openrouter.ai/api/v1",
  tokenlayKey: process.env.TOKENLAY_KEY,
});

Supported Models

• All models available on OpenRouter platform
• Dynamic model routing

Multi-Provider Setup

Configure multiple providers in a single client:

const client = new TokenlayOpenAI({
  openai_key: process.env.OPENAI_API_KEY,
  anthropic_key: process.env.ANTHROPIC_API_KEY,
  google_key: process.env.GOOGLE_API_KEY,
  tokenlayKey: process.env.TOKENLAY_KEY,
});

Tokenlay will automatically route requests to the appropriate provider based on the model specified in your API calls.

Provider Detection Summary

Provider-Specific Notes

AWS Bedrock

• Requires AWS credentials (access key, secret key, region)
• Models use Bedrock-specific identifiers
• Additional IAM permissions may be required

Azure OpenAI

• Requires custom base URL for your Azure resource
• Model names match your Azure deployment names
• No automatic key detection due to custom format

OpenRouter

• Provides access to hundreds of models from various providers
• Uses OpenAI-compatible API format
• Requires explicit base URL configuration