# Claude (Anthropic) Compatibility Guide > **Important Note**: We provide API services compatible with Anthropic standards. Therefore, we strongly recommend directly referring to the [official Anthropic API documentation](https://docs.anthropic.com/en/api/messages) to obtain the most comprehensive, up-to-date parameter details and examples. The following is an overview of the simplified interface, focusing on core fields and usage instructions. ## Overview UModelverse platform provides a **Messages** interface compatible with the Anthropic Claude API, allowing developers to directly invoke models on Modelverse using the Anthropic SDK or other supported tools. - **Request Method**: POST - **Endpoint**: `https://api-us-ca.umodelverse.ai/v1/messages` - **Authentication**: Use API key, pass through `x-api-key: {api_key}` or `Authorization: Bearer {api_key}` > **Note**: The `/v1/messages` endpoint only supports the Claude series models. For other models (such as GPT, Gemini, DeepSeek, etc.), use the OpenAI-compatible `/v1/chat/completions` endpoint. Although Claude and OpenAI both use `/v1/models` to obtain a list of models, the call interfaces are different. ## Key Core Fields ### Request Fields | Field | Type | Required | Default | Description | | -------------- | ------------ | -------- | ------- | --------------------------------------------------------------------------- | | model | string | Yes | None | Model ID, such as `claude-sonnet-4-5-20250929`. Specifies the model for generating responses. | | messages | array | Yes | None | List of conversation messages. Each message includes `role` (user/assistant) and `content` (text/image). | | max_tokens | integer | Yes | None | Maximum number of tokens to generate. Controls the response length to prevent overly lengthy outputs. | | system | string/array | No | None | System prompts. Defines model behavior and roles. | | temperature | number | No | 1.0 | Sampling temperature (0 to 1). Controls randomness, with higher values being more creative and lower values being more deterministic. | | top_p | number | No | None | Nucleus sampling (0 to 1). Controls diversity. | | top_k | integer | No | None | Top-K sampling. Samples only from the top K tokens with the highest probabilities. | | stop_sequences | array | No | None | List of stop sequences. Stops generation upon reaching these sequences. | | stream | boolean | No | false | Whether to stream responses. Returns content incrementally in real-time. | | metadata | object | No | None | Metadata object, which can include `user_id` for tracking. | | tools | array | No | None | List of available tools. Enables function calling features. | | tool_choice | object | No | auto | Tool selection strategy. For example, `{"type": "auto"}` lets the model decide to invoke tools. | ### Response Fields | Field | Type | Description | | ------------- | ------ | --------------------------------------------------------------- | | id | string | Response ID. Uniquely identifies this completion. | | type | string | Object type: `message`. | | role | string | Role: `assistant`. | | content | array | List of content blocks. Each block contains `type` and corresponding content (such as `text`). | | model | string | The model ID used. | | stop_reason | string | Stop reason: `end_turn`, `max_tokens`, `stop_sequence`, `tool_use`. | | stop_sequence | string | Sequence that triggered the stop (if applicable). | | usage | object | Usage statistics. Includes `input_tokens` and `output_tokens`. | ## Quick Start ### Install Anthropic SDK ```bash pip install anthropic ``` ### Sample Code #### ** Python ** ```python import anthropic client = anthropic.Anthropic( api_key="", base_url="https://api-us-ca.umodelverse.ai" ) message = client.messages.create( model="claude-sonnet-4-5-20250929", max_tokens=1024, messages=[ {"role": "user", "content": "Hello, how are you?"} ] ) print(message.content[0].text) ``` #### ** Python Streaming ** ```python import anthropic client = anthropic.Anthropic( api_key="", base_url="https://api-us-ca.umodelverse.ai" ) with client.messages.stream( model="claude-sonnet-4-5-20250929", max_tokens=1024, messages=[ {"role": "user", "content": "Write a short story about AI."} ] ) as stream: for text in stream.text_stream: print(text, end="", flush=True) ``` #### ** curl ** ```bash curl https://api-us-ca.umodelverse.ai/v1/messages \ -H "Content-Type: application/json" \ -H "x-api-key: $MODELVERSE_API_KEY" \ -H "anthropic-version: 2023-06-01" \ -d '{ "model": "claude-sonnet-4-5-20250929", "max_tokens": 1024, "messages": [ {"role": "user", "content": "Hello, how are you?"} ] }' ``` #### ** curl Streaming ** ```bash curl https://api-us-ca.umodelverse.ai/v1/messages \ -H "Content-Type: application/json" \ -H "x-api-key: $MODELVERSE_API_KEY" \ -H "anthropic-version: 2023-06-01" \ -d '{ "model": "claude-sonnet-4-5-20250929", "max_tokens": 1024, "stream": true, "messages": [ {"role": "user", "content": "Hello, how are you?"} ] }' ``` ## Multimodal Support Claude-compatible interfaces support image input, which can be delivered via base64 encoding or URL: ### Using URL ```python import anthropic client = anthropic.Anthropic( api_key="", base_url="https://api-us-ca.umodelverse.ai" ) # Using URL to pass image message = client.messages.create( model="claude-sonnet-4-5-20250929", max_tokens=1024, messages=[ { "role": "user", "content": [ { "type": "image", "source": { "type": "url", "url": "https://umodelverse-inference.cn-wlcb.ufileos.com/ucloud-maxcot.jpg" } }, { "type": "text", "text": "What's in this image?" } ] } ] ) print(message.content[0].text) ``` ### Using base64 Encoding ```python import anthropic import base64 client = anthropic.Anthropic( api_key="", base_url="https://api-us-ca.umodelverse.ai" ) # Using base64 to encode image with open("image.png", "rb") as f: image_data = base64.standard_b64encode(f.read()).decode("utf-8") message = client.messages.create( model="claude-sonnet-4-5-20250929", max_tokens=1024, messages=[ { "role": "user", "content": [ { "type": "image", "source": { "type": "base64", "media_type": "image/png", "data": image_data } }, { "type": "text", "text": "What's in this image?" } ] } ] ) print(message.content[0].text) ``` ## Tool Calling (Function Calling) ```python import anthropic client = anthropic.Anthropic( api_key="", base_url="https://api-us-ca.umodelverse.ai" ) tools = [ { "name": "get_weather", "description": "Get the current weather in a given location", "input_schema": { "type": "object", "properties": { "location": { "type": "string", "description": "The city and state, e.g. San Francisco, CA" } }, "required": ["location"] } } ] message = client.messages.create( model="claude-sonnet-4-5-20250929", max_tokens=1024, tools=tools, messages=[ {"role": "user", "content": "What's the weather like in Beijing?"} ] ) print(message.content) ``` ## Model List For more supported Claude-compatible models, please see [Get Model List] or visit the [UModelverse Model Center](https://console.ucloud.cn/modelverse/model-center). > For more details on the fields, see the [official Anthropic documentation](https://docs.anthropic.com/en/api/messages)