Anthropic Adapter

The Anthropic adapter translates Tekimax's unified interface to Anthropic's Messages API. Under the hood it uses the official @anthropic-ai/sdk package.

Installation

Code

npm install tekimax-ts

Usage

Code

import { Tekimax, AnthropicProvider } from 'tekimax-ts';

const client = new Tekimax({
  provider: new AnthropicProvider({
    apiKey: process.env.ANTHROPIC_API_KEY!,
  })
});

const result = await client.text.chat.completions.create({
  model: 'claude-3-5-sonnet-20240620',
  messages: [{ role: 'user', content: 'Hello Claude' }]
});

console.log(result.message.content);

Streaming

Code

const stream = client.text.chat.completions.createStream({
  model: 'claude-3-5-sonnet-20240620',
  messages: [{ role: 'user', content: 'Tell me a story' }]
});

for await (const chunk of stream) {
  process.stdout.write(chunk.delta);
}

Tool Calling

Tool calling works identically to OpenAI — the adapter handles the translation to Anthropic's tool_use content blocks behind the scenes.

Code

const result = await client.text.chat.completions.create({
  model: 'claude-3-5-sonnet-20240620',
  messages: [{ role: 'user', content: 'What is the weather in Paris?' }],
  tools: [{
    type: 'function',
    function: {
      name: 'get_weather',
      description: 'Get current weather for a location',
      parameters: {
        type: 'object',
        properties: { location: { type: 'string' } },
        required: ['location']
      }
    }
  }]
});

// Normalized to the same shape as every other provider.
console.log(result.message.toolCalls);

Vision (Image Analysis)

Code

const analysis = await client.images.analyze({
  model: 'claude-3-5-sonnet-20240620',
  image: 'https://example.com/photo.jpg', // URL or Base64 data URI
  prompt: 'Describe what you see in this image.'
});

console.log(analysis.content);

Notes

System Prompts: Anthropic's Messages API requires system as a top-level parameter, not a message role. The adapter automatically extracts any { role: 'system' } messages from your array and passes them to the system field — so you can use the same message format as every other provider.
Max Tokens: Anthropic requires a max_tokens value on every request. The adapter defaults to 1024 if you don't set maxTokens, so you don't need to remember this requirement.
Tool Result Mapping: Anthropic expects tool results as a user message with tool_result content blocks, not as a role: 'tool' message. The adapter handles this conversion automatically.

Anthropic Adapter

Installation

Usage

Streaming

Tool Calling

Vision (Image Analysis)

Notes

On this page