你当前正在访问 Microsoft Azure Global Edition 技术文档网站。 如果需要访问由世纪互联运营的 Microsoft Azure 中国技术文档网站,请访问 https://docs.azure.cn

Azure AI 代理服务支持的模型

代理由一组具有不同功能和价位的模型提供支持。 模型可用性因区域和云而异。 某些工具和功能需要最新的模型。 可用 SDK 中提供了以下模型。 下表适用于即用即付。 有关预配吞吐量单位 (PTU) 可用性的信息,请参阅 Azure OpenAI 文档中预配的吞吐量。 如果此处列出的区域支持全局标准模型,则可以使用全局标准模型

Azure OpenAI 模型

Azure AI 代理服务支持与 Azure OpenAI 中的聊天完成 API 相同的模型,这些模型位于以下区域。

区域 gpt-4o,2024-05-13 gpt-4o,2024-08-06 gpt-4o-mini,2024-07-18 gpt-40613 gpt-41106-Preview gpt-40125-Preview gpt-4,turbo-2024-04-09 gpt-4-32k0613 gpt-35-turbo0613 gpt-35-turbo1106 gpt-35-turbo0125 gpt-35-turbo-16k0613
australiaeast - - - - -
eastus - - - -
eastus2 - - - -
francecentral - - - - - -
日本东部 - - - - - - - - -
norwayeast - - - - - - - - - - -
southindia - - - - - - - - -
瑞典中部 - -
uksouth - - - - - -
westus - - - - -
westus3 - - - - - -

非 Microsoft 模型

Azure AI 代理服务还支持 Azure AI Foundry 模型目录中的以下模型。

  • Meta-Llama-405B-Instruct
  • Mistral-large-2407
  • Cohere-command-r-plus
  • Cohere-command-r

若要使用这些模型,可以使用 Azure AI Foundry 门户进行部署,然后在代理中引用部署名称。 例如:

agent = project_client.agents.create_agent( model="llama-3", name="my-agent", instructions="You are a helpful agent" ) 

后续步骤

创建新的代理项目