1# Compatible Anthropic API
2 
3> Call MiniMax models using the Anthropic SDK
4 
5To meet developers' needs for the Anthropic API ecosystem, our API now supports the Anthropic API format. With simple configuration, you can integrate MiniMax capabilities into the Anthropic API ecosystem.
6 
7## Quick Start
8 
9### 1. Install Anthropic SDK
10 
11<CodeGroup>
12  ```bash Python theme={null}
13  pip install anthropic
14  ```
15 
16  ```bash Node.js theme={null}
17  npm install @anthropic-ai/sdk
18  ```
19</CodeGroup>
20 
21### 2. Configure Environment Variables
22 
23For international users, use `https://api.minimax.io/anthropic`; for users in China, use `https://api.minimaxi.com/anthropic`
24 
25```bash  theme={null}
26export ANTHROPIC_BASE_URL=https://api.minimax.io/anthropic
27export ANTHROPIC_API_KEY=${YOUR_API_KEY}
28```
29 
30### 3. Call API
31 
32```python Python theme={null}
33import anthropic
34 
35client = anthropic.Anthropic()
36 
37message = client.messages.create(
38    model="MiniMax-M2.1",
39    max_tokens=1000,
40    system="You are a helpful assistant.",
41    messages=[
42        {
43            "role": "user",
44            "content": [
45                {
46                    "type": "text",
47                    "text": "Hi, how are you?"
48                }
49            ]
50        }
51    ]
52)
53 
54for block in message.content:
55    if block.type == "thinking":
56        print(f"Thinking:\n{block.thinking}\n")
57    elif block.type == "text":
58        print(f"Text:\n{block.text}\n")
59```
60 
61### 4. Important Note
62 
63In multi-turn function call conversations, the complete model response (i.e., the assistant message) must be append to the conversation history to maintain the continuity of the reasoning chain.
64 
65* Append the full `response.content` list to the message history (includes all content blocks: thinking/text/tool\_use)
66 
67## Supported Models
68 
69When using the Anthropic SDK, the `MiniMax-M2.1` `MiniMax-M2.1-lightning` `MiniMax-M2` model is supported:
70 
71| Model Name             | Description                                                                                                                               |
72| :--------------------- | :---------------------------------------------------------------------------------------------------------------------------------------- |
73| MiniMax-M2.1           | Powerful Multi-Language Programming Capabilities with Comprehensively Enhanced Programming Experience (output speed approximately 60 tps) |
74| MiniMax-M2.1-lightning | Faster and More Agile (output speed approximately 100 tps)                                                                                |
75| MiniMax-M2             | Agentic capabilities, Advanced reasoning                                                                                                  |
76 
77<Note>
78  The Anthropic API compatibility interface currently only supports the
79  `MiniMax-M2.1` `MiniMax-M2.1-lightning` `MiniMax-M2` model. For other models, please use the standard MiniMax API
80  interface.
81</Note>
82 
83## Compatibility
84 
85### Supported Parameters
86 
87When using the Anthropic SDK, we support the following input parameters:
88 
89| Parameter            | Support Status  | Description                                                         |
90| :------------------- | :-------------- | :------------------------------------------------------------------ |
91| `model`              | Fully supported | supports `MiniMax-M2.1` `MiniMax-M2.1-lightning` `MiniMax-M2` model |
92| `messages`           | Partial support | Supports text and tool calls, no image/document input               |
93| `max_tokens`         | Fully supported | Maximum number of tokens to generate                                |
94| `stream`             | Fully supported | Streaming response                                                  |
95| `system`             | Fully supported | System prompt                                                       |
96| `temperature`        | Fully supported | Range (0.0, 1.0], controls output randomness, recommended value: 1  |
97| `tool_choice`        | Fully supported | Tool selection strategy                                             |
98| `tools`              | Fully supported | Tool definitions                                                    |
99| `top_p`              | Fully supported | Nucleus sampling parameter                                          |
100| `metadata`           | Fully Supported | Metadata                                                            |
101| `thinking`           | Fully Supported | Reasoning Content                                                   |
102| `top_k`              | Ignored         | This parameter will be ignored                                      |
103| `stop_sequences`     | Ignored         | This parameter will be ignored                                      |
104| `service_tier`       | Ignored         | This parameter will be ignored                                      |
105| `mcp_servers`        | Ignored         | This parameter will be ignored                                      |
106| `context_management` | Ignored         | This parameter will be ignored                                      |
107| `container`          | Ignored         | This parameter will be ignored                                      |
108 
109### Messages Field Support
110 
111| Field Type           | Support Status  | Description                      |
112| :------------------- | :-------------- | :------------------------------- |
113| `type="text"`        | Fully supported | Text messages                    |
114| `type="tool_use"`    | Fully supported | Tool calls                       |
115| `type="tool_result"` | Fully supported | Tool call results                |
116| `type="thinking"`    | Fully supported | Reasoning Content                |
117| `type="image"`       | Not supported   | Image input not supported yet    |
118| `type="document"`    | Not supported   | Document input not supported yet |
119 
120## Examples
121 
122### Streaming Response
123 
124```python Python theme={null}
125import anthropic
126 
127client = anthropic.Anthropic()
128 
129print("Starting stream response...\n")
130print("=" * 60)
131print("Thinking Process:")
132print("=" * 60)
133 
134stream = client.messages.create(
135    model="MiniMax-M2.1",
136    max_tokens=1000,
137    system="You are a helpful assistant.",
138    messages=[
139        {"role": "user", "content": [{"type": "text", "text": "Hi, how are you?"}]}
140    ],
141    stream=True,
142)
143 
144reasoning_buffer = ""
145text_buffer = ""
146 
147for chunk in stream:
148    if chunk.type == "content_block_start":
149        if hasattr(chunk, "content_block") and chunk.content_block:
150            if chunk.content_block.type == "text":
151                print("\n" + "=" * 60)
152                print("Response Content:")
153                print("=" * 60)
154 
155    elif chunk.type == "content_block_delta":
156        if hasattr(chunk, "delta") and chunk.delta:
157            if chunk.delta.type == "thinking_delta":
158                # Stream output thinking process
159                new_thinking = chunk.delta.thinking
160                if new_thinking:
161                    print(new_thinking, end="", flush=True)
162                    reasoning_buffer += new_thinking
163            elif chunk.delta.type == "text_delta":
164                # Stream output text content
165                new_text = chunk.delta.text
166                if new_text:
167                    print(new_text, end="", flush=True)
168                    text_buffer += new_text
169 
170print("\n")
171```
172 
173### Tool Use & Interleaved Thinking
174 
175Learn how to use M2.1 Tool Use and Interleaved Thinking capabilities with Anthropic SDK, please refer to the following documentation.
176 
177<Columns cols={1}>
178  <Card title="M2.1 Tool Use & Interleaved Thinking" icon="book-open" href="/guides/text-m2-function-call#anthropic-sdk" arrow="true" cta="Click here">
179    Learn how to leverage MiniMax-M2.1 tool calling and interleaved thinking capabilities to enhance performance in complex tasks.
180  </Card>
181</Columns>
182 
183## Important Notes
184 
185<Warning>
186  1. The Anthropic API compatibility interface currently only supports the `MiniMax-M2.1` `MiniMax-M2` model
187 
188  2. The `temperature` parameter range is (0.0, 1.0], values outside this range will return an error
189 
190  3. Some Anthropic parameters (such as `thinking`, `top_k`, `stop_sequences`, `service_tier`, `mcp_servers`, `context_management`, `container`) will be ignored
191 
192  4. Image and document type inputs are not currently supported
193</Warning>
194 
195## Related Links
196 
197* [Anthropic SDK Documentation](https://docs.anthropic.com/en/api/client-sdks)
198* [MiniMax Text Generation API](/api-reference/text-intro)
199* [M2.1 Tool Use & Interleaved Thinking](/guides/text-m2-function-call)
200 
201## Recommended Reading
202 
203<Columns cols={2}>
204  <Card title="Text Generation" icon="book-open" href="/guides/text-generation" arrow="true" cta="Click here">
205    Supports text generation via compatible Anthropic API and OpenAI API.
206  </Card>
207 
208  <Card title="Compatible OpenAI API" icon="book-open" href="/api-reference/text-openai-api" arrow="true" cta="Click here">
209    Use OpenAI SDK with MiniMax models
210  </Card>
211 
212  <Card title="M2.1 for AI Coding Tools" icon="book-open" href="/guides/text-ai-coding-tools" arrow="true" cta="Click here">
213    MiniMax-M2.1 excels at code understanding, dialogue, and reasoning.
214  </Card>
215 
216  <Card title="M2.1 Tool Use & Interleaved Thinking" icon="book-open" href="/guides/text-m2-function-call" arrow="true" cta="Click here">
217    AI models can call external functions to extend their capabilities.
218  </Card>
219</Columns>
220 
221 
222---
223 
224> To find navigation and other pages in this documentation, fetch the llms.txt file at: https://platform.minimax.io/docs/llms.txt