兼容 OpenAI SDK

为方便您集成和使用，火山方舟模型 API 会尽可能兼容 OpenAI API。修改 base url、model 、api_key等少量配置，轻松将方舟模型服务集成到您已有的系统中。

前提条件#

OpenAI SDK#

Python版本：3.7及以上。
OpenAI SDK：1.0版本及以上，安装命令：
Python
```
1pip install --upgrade "openai>=1.0"
```

快速开始示例#

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18from openai import OpenAI
import os

client = OpenAI(
    # The base URL for model invocation
    base_url="https://ark.cn-beijing.volces.com/api/v3",
    # Replace with your API Key
    api_key=os.environ.get("ARK_API_KEY"),
)

completion = client.chat.completions.create(
    # Replace with Model ID
    model="doubao-seed-2-1-pro-260915",
    messages = [
        {"role": "user", "content": "Hello"},
    ],
)
print(completion.choices[0].message.content)

设置额外字段#

传入OpenAI SDK中不支持的字段，可以通过 extra_body 字典传入，如开关模型是否深度思考的 thinking 字段。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24from openai import OpenAI
import os

client = OpenAI(
    # The base URL for model invocation
    base_url="https://ark.cn-beijing.volces.com/api/v3",
    # Replace with your API Key
    api_key=os.environ.get("ARK_API_KEY"),
)

completion = client.chat.completions.create(
    # Replace with Model ID
    model="doubao-seed-2-1-pro-260915",
    messages = [
        {"role": "user", "content": "Hello"},
    ],
    extra_body={
         "thinking": {
             "type": "disabled", # 不使用深度思考能力
             # "type": "enabled", # 使用深度思考能力
         }
     }
)
print(completion.choices[0].message.content)

设置自定义header#

可以用于传递额外信息，如配置 ID来串联日志，使能数据加密能力。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20from openai import OpenAI
import os

client = OpenAI(
    # The base URL for model invocation
    base_url="https://ark.cn-beijing.volces.com/api/v3",
    # Replace with your API Key
    api_key=os.environ.get("ARK_API_KEY"),
)

completion = client.chat.completions.create(
    # Replace with Model ID
    model="doubao-seed-2-1-pro-260915",
    messages = [
        {"role": "user", "content": "Hello"},
    ],
    # 自定义request id
    extra_headers={"X-Client-Request-Id": "202406251728190000B7EA7A9648AC08D9"}
)
print(completion.choices[0].message.content)

向量化 Embedding#

LangChain OpenAI SDK#

安装 LangChain OpenAI SDK：

1pip install langchain-openai

示例代码#

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24from langchain_openai import ChatOpenAI
from langchain_core.prompts import PromptTemplate
import os

llm = ChatOpenAI(
    # Replace with your API Key
    openai_api_key=os.environ.get("ARK_API_KEY"),
    # The base URL for model invocation
    openai_api_base="https://ark.cn-beijing.volces.com/api/v3",
    # Replace with Model ID
    model="doubao-seed-2-1-pro-260915",
)

template = """Question: {question}

Answer: Let's think step by step."""

prompt = PromptTemplate.from_template(template)

question = "What NFL team won the Super Bowl in the year Justin Beiber was born?"

llm_chain = prompt | llm

print(llm_chain.invoke(question))