Anthropic SDK Integration

CloudBase AI is compatible with both OpenAI Chat Completions and Anthropic Messages API protocols. You can use the Anthropic SDK directly by passing authToken and baseURL in the client constructor.

Online Example

Open full example code in CodeSandbox →

Authentication

CloudBase authenticates via Authorization: Bearer <token>, which corresponds to the authToken parameter in the Anthropic SDK — not apiKey.

Prerequisites

1. A CloudBase environment (older plans can be upgraded), with model switch enabled (see Overview)
2. An API Key is created (Get one here)

Installation

# Node.js
npm install @anthropic-ai/sdk

# Python
pip install anthropic

Configuration

Pass authToken and baseURL when initializing the client:

Parameter	Value
authToken	Your CloudBase API Key
baseURL	https://<ENV_ID>.api.tcloudbasegateway.com/v1/ai/cloudbase

Replace <ENV_ID> with your actual CloudBase environment ID and <YOUR_CLOUDBASE_API_KEY> with your actual API Key.

Node.js Examples

Non-streaming

const Anthropic = require("@anthropic-ai/sdk");

const client = new Anthropic({
  authToken: "<YOUR_CLOUDBASE_API_KEY>",
  baseURL: "https://<ENV_ID>.api.tcloudbasegateway.com/v1/ai/cloudbase"
});

async function main() {
  const message = await client.messages.create({
    model: "hy3-preview",
    max_tokens: 1024,
    messages: [{ role: "user", content: "Hello" }]
  });

  console.log(message.content[0].text);
}

main();

Streaming

const Anthropic = require("@anthropic-ai/sdk");

const client = new Anthropic({
  authToken: "<YOUR_CLOUDBASE_API_KEY>",
  baseURL: "https://<ENV_ID>.api.tcloudbasegateway.com/v1/ai/cloudbase"
});

async function main() {
  const stream = await client.messages.stream({
    model: "hy3-preview",
    max_tokens: 1024,
    messages: [{ role: "user", content: "Introduce yourself" }]
  });

  for await (const chunk of stream) {
    if (
      chunk.type === "content_block_delta" &&
      chunk.delta.type === "text_delta"
    ) {
      process.stdout.write(chunk.delta.text);
    }
  }
}

main();

Python Examples

Non-streaming

import anthropic

client = anthropic.Anthropic(
    auth_token="<YOUR_CLOUDBASE_API_KEY>",
    base_url="https://<ENV_ID>.api.tcloudbasegateway.com/v1/ai/cloudbase"
)

message = client.messages.create(
    model="hy3-preview",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello"}]
)

print(message.content[0].text)

Streaming

import anthropic

client = anthropic.Anthropic(
    auth_token="<YOUR_CLOUDBASE_API_KEY>",
    base_url="https://<ENV_ID>.api.tcloudbasegateway.com/v1/ai/cloudbase"
)

with client.messages.stream(
    model="hy3-preview",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Introduce yourself"}]
) as stream:
    for text in stream.text_stream:
        print(text, end="", flush=True)

Multi-turn Conversations

Multi-turn conversations require maintaining a complete messages array in each request. See the Multi-turn Conversation documentation for details.

Migrating from Anthropic

If your project already uses the Anthropic SDK, update the client initialization to use authToken instead of apiKey:

// Original Anthropic configuration
const client = new Anthropic({
  apiKey: "<ANTHROPIC_API_KEY>"
});

// CloudBase configuration
const client = new Anthropic({
  authToken: "<YOUR_CLOUDBASE_API_KEY>",
  baseURL: "https://<ENV_ID>.api.tcloudbasegateway.com/v1/ai/cloudbase"
});

Replace the model name with a model enabled in CloudBase (e.g., hy3-preview). No other code changes are needed.

Error Handling

const Anthropic = require("@anthropic-ai/sdk");

const client = new Anthropic({
  authToken: "<YOUR_CLOUDBASE_API_KEY>",
  baseURL: "https://<ENV_ID>.api.tcloudbasegateway.com/v1/ai/cloudbase"
});

try {
  const message = await client.messages.create({
    model: "hy3-preview",
    max_tokens: 1024,
    messages: [{ role: "user", content: "Hello" }]
  });
  console.log(message.content[0].text);
} catch (error) {
  if (error instanceof Anthropic.APIError) {
    console.error("API Error:", error.status, error.message);
  } else {
    throw error;
  }
}

Supported Parameters

Parameter	Type	Description
model	string	Model name
messages	array	List of messages
system	string	System prompt
max_tokens	number	Maximum tokens to generate (required)
stream	boolean	Enable streaming
temperature	number	Sampling temperature (0–1)
top_p	number	Nucleus sampling (0–1)