Provider Key Inference

Tokenlay can automatically detect AI providers from your API key prefixes, making configuration seamless across multiple providers.

Automatic Detection

When you use provider_api_key, Tokenlay attempts to infer the provider based on the key prefix:

const client = new TokenlayOpenAI({
  provider_api_key: "sk-...", // Detected as OpenAI
  tokenlayKey: process.env.TOKENLAY_KEY,
});

const client = new TokenlayOpenAI({
  provider_api_key: "sk-ant-...", // Detected as Anthropic
  tokenlayKey: process.env.TOKENLAY_KEY,
});

Explicit Provider Keys

For explicit control, use provider-specific key parameters:

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

Key Prefix Patterns

Provider	Key Prefix	Parameter Name
OpenAI	`sk-`	`openai_key`
Anthropic	`sk-ant-`	`anthropic_key`
Google	`AIza`	`google_key`
AWS Bedrock	`AKIA`	`aws_access_key`
Azure OpenAI	Custom	`azure_key`

Mixed Provider Setup

You can 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 route requests to the appropriate provider based on the model specified in your API calls.

Advanced Configuration Base URL Configuration