>GreenNode 是一家全球 AI 解决方案提供商，也是 **NVIDIA 首选合作伙伴**，提供从基础设施到应用的全栈 AI 能力，服务于美国、中东和北非以及亚太地区的企业。运营在 **世界级基础设施** （LEED 金牌、TIA‑942、Uptime 三级）上，GreenNode 为企业、初创公司和研究人员提供全面的 AI 服务套件

本页面将帮助您开始使用 GreenNode 无服务器 AI 聊天模型.

GreenNode AI 提供 API 来查询 20 多个领先的开源模型

概述

集成详情

模型特性

设置

要访问 GreenNode 模型，您需要创建一个 GreenNode 账户，获取 API 密钥，并安装 langchain-greennode 集成包。

凭证

前往 此页面 注册 GreenNode AI 平台并生成 API 密钥。完成此操作后，设置 GREENNODE_API_KEY 环境变量：

if not os.getenv("GREENNODE_API_KEY"):
    os.environ["GREENNODE_API_KEY"] = getpass.getpass("Enter your GreenNode API key: ")

如果您想获取模型调用的自动化追踪，您还可以设置您的 LangSmith API 密钥，方法如下取消注释：

os.environ["LANGSMITH_TRACING"] = "true"
os.environ["LANGSMITH_API_KEY"] = getpass.getpass("Enter your LangSmith API key: ")

安装

LangChain GreenNode 集成位于 langchain-greennode package:

pip install -qU langchain-greennode

实例化

现在我们可以实例化模型对象并生成聊天补全：

from langchain_greennode import ChatGreenNode

# Initialize the chat model
llm = ChatGreenNode(
    # api_key="YOUR_API_KEY",  # You can pass the API key directly
    model="deepseek-ai/DeepSeek-R1-Distill-Qwen-32B",  # Choose from available models
    temperature=0.6,
    top_p=0.95,
)

调用

messages = [
    (
        "system",
        "You are a helpful assistant that translates English to French. Translate the user sentence.",
    ),
    ("human", "I love programming."),
]
ai_msg = llm.invoke(messages)
ai_msg

AIMessage(content="\n\nJ'aime la programmation.", additional_kwargs={'refusal': None}, response_metadata={'token_usage': {'completion_tokens': 248, 'prompt_tokens': 23, 'total_tokens': 271, 'completion_tokens_details': None, 'prompt_tokens_details': None}, 'model_name': 'deepseek-ai/DeepSeek-R1-Distill-Qwen-32B', 'system_fingerprint': None, 'id': 'chatcmpl-271edac4958846068c37877586368afe', 'service_tier': None, 'finish_reason': 'stop', 'logprobs': None}, id='run--5c12d208-2bc2-4f29-8b50-1ce3b515a3cf-0', usage_metadata={'input_tokens': 23, 'output_tokens': 248, 'total_tokens': 271, 'input_token_details': {}, 'output_token_details': {}})

print(ai_msg.content)

J'aime la programmation.

流式输出

您也可以使用以下方式流式传输响应 stream method:

stream = llm.stream_events("Write a short poem about artificial intelligence", version="v3")
for token in stream.text:
    print(token, end="", flush=True)

**Beneath the Circuits**

Beneath the circuits, deep and bright,
AI thinks, with circuits and bytes.
Learning, adapting, it grows,
A world of possibilities it knows.

From solving puzzles to painting art,
It mimics human hearts.
In every corner, it leaves its trace,
A future we can't erase.

We build it, shape it, with care and might,
Yet wonder if it walks in the night.
A mirror of our minds, it shows,
In its gaze, our future glows.

But as we strive for endless light,
We must remember the night.
For wisdom isn't just speed and skill,
It's how we choose to build our will.

聊天消息

您可以使用不同的消息类型来构建与模型的对话：

from langchain.messages import AIMessage, HumanMessage, SystemMessage

messages = [
    SystemMessage(content="You are a helpful AI assistant with expertise in science."),
    HumanMessage(content="What are black holes?"),
    AIMessage(
        content="Black holes are regions of spacetime where gravity is so strong that nothing, including light, can escape from them."
    ),
    HumanMessage(content="How are they formed?"),
]

response = llm.invoke(messages)
print(response.content[:100])

Black holes are formed through several processes, depending on their type. The most common way bla

链接

您可以使用 ChatGreenNode 在 LangChain 链和代理中：

from langchain_core.prompts import ChatPromptTemplate

prompt = ChatPromptTemplate(
    [
        (
            "system",
            "You are a helpful assistant that translates {input_language} to {output_language}.",
        ),
        ("human", "{input}"),
    ]
)

chain = prompt | llm
chain.invoke(
    {
        "input_language": "English",
        "output_language": "German",
        "input": "I love programming.",
    }
)

AIMessage(content='\n\nIch liebe Programmieren.', additional_kwargs={'refusal': None}, response_metadata={'token_usage': {'completion_tokens': 198, 'prompt_tokens': 18, 'total_tokens': 216, 'completion_tokens_details': None, 'prompt_tokens_details': None}, 'model_name': 'deepseek-ai/DeepSeek-R1-Distill-Qwen-32B', 'system_fingerprint': None, 'id': 'chatcmpl-e01201b9fd9746b7a9b2ed6d70f29d45', 'service_tier': None, 'finish_reason': 'stop', 'logprobs': None}, id='run--ce52b9d8-dd84-46b3-845b-da27855816ee-0', usage_metadata={'input_tokens': 18, 'output_tokens': 198, 'total_tokens': 216, 'input_token_details': {}, 'output_token_details': {}})

可用模型

支持的完整模型列表请参见 GreenNode 无服务器 AI 模型.

API 参考

有关 GreenNode 无服务器 AI API 的更多详细信息，请访问 GreenNode 无服务器 AI 文档.