Azure OpenAI 推理模型

Azure OpenAI 推理模型旨在处理推理任务和解决问题的任务，具有更好的针对性和功能。 与之前的迭代相比，这些模型花费更多的时间处理和理解用户的请求，使它们在科学、编码和数学等领域非常强大。

推理模型的主要功能：

• 复杂代码生成：能够生成算法并处理高级编码任务以支持开发人员。
• 高级问题解决：非常适合全面的头脑风暴会议和解决多方面的挑战。
• 复杂文档比较：非常适合分析合同、案例文件或法律文档以识别细微的差别。
• 指令遵循和工作流管理：对于管理需要较短上下文的工作流特别有效。

可用性

区域可用性

API 和功能支持

不支持

推理模型当前不支持以下各项：

• temperature、top_p、presence_penalty、frequency_penalty、logprobs、top_logprobs、logit_bias、max_tokens

用法

这些模型 当前不支持与 使用聊天完成 API 的其他模型相同的参数集。

pip install openai --upgrade

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.getenv("AZURE_OPENAI_API_KEY"),
    base_url="https://YOUR-RESOURCE-NAME.openai.azure.com/openai/v1/",
)

response = client.chat.completions.create(
    model="gpt-5-mini", # replace with the model deployment name of your o1 deployment.
    messages=[
        {"role": "user", "content": "What steps should I think about when writing my first Python API?"},
    ],
    max_completion_tokens = 5000

)

print(response.model_dump_json(indent=2))

pip install openai --upgrade

from openai import OpenAI
from azure.identity import DefaultAzureCredential, get_bearer_token_provider

token_provider = get_bearer_token_provider(
    DefaultAzureCredential(), "https://cognitiveservices.azure.com/.default"
)

client = OpenAI(  
  base_url = "https://YOUR-RESOURCE-NAME.openai.azure.com/openai/v1/",  
  api_key=token_provider,
)

response = client.chat.completions.create(
    model="o1-new", # replace with your model deployment name 
    messages=[
        {"role": "user", "content": "What steps should I think about when writing my first Python API?"},
    ],
    max_completion_tokens = 5000

)

print(response.model_dump_json(indent=2))

using Azure.Identity;
using OpenAI;
using OpenAI.Chat;
using System.ClientModel.Primitives;

#pragma warning disable OPENAI001 //currently required for token based authentication

BearerTokenPolicy tokenPolicy = new(
    new DefaultAzureCredential(),
    "https://cognitiveservices.azure.com/.default");

ChatClient client = new(
    model: "o4-mini",
    authenticationPolicy: tokenPolicy,
    options: new OpenAIClientOptions()
    {

        Endpoint = new Uri("https://YOUR-RESOURCE-NAME.openai.azure.com/openai/v1")
    }
);

ChatCompletionOptions options = new ChatCompletionOptions
{
    MaxOutputTokenCount = 100000
};

ChatCompletion completion = client.CompleteChat(
         new DeveloperChatMessage("You are a helpful assistant"),
         new UserChatMessage("Tell me about the bitter lesson")
    );

Console.WriteLine($"[ASSISTANT]: {completion.Content[0].Text}");

Python 输出：

{
  "id": "chatcmpl-AEj7pKFoiTqDPHuxOcirA9KIvf3yz",
  "choices": [
    {
      "finish_reason": "stop",
      "index": 0,
      "logprobs": null,
      "message": {
        "content": "Writing your first Python API is an exciting step in developing software that can communicate with other applications. An API (Application Programming Interface) allows different software systems to interact with each other, enabling data exchange and functionality sharing. Here are the steps you should consider when creating your first Python API...truncated for brevity.",
        "refusal": null,
        "role": "assistant",
        "function_call": null,
        "tool_calls": null
      },
      "content_filter_results": {
        "hate": {
          "filtered": false,
          "severity": "safe"
        },
        "protected_material_code": {
          "filtered": false,
          "detected": false
        },
        "protected_material_text": {
          "filtered": false,
          "detected": false
        },
        "self_harm": {
          "filtered": false,
          "severity": "safe"
        },
        "sexual": {
          "filtered": false,
          "severity": "safe"
        },
        "violence": {
          "filtered": false,
          "severity": "safe"
        }
      }
    }
  ],
  "created": 1728073417,
  "model": "o1-2024-12-17",
  "object": "chat.completion",
  "service_tier": null,
  "system_fingerprint": "fp_503a95a7d8",
  "usage": {
    "completion_tokens": 1843,
    "prompt_tokens": 20,
    "total_tokens": 1863,
    "completion_tokens_details": {
      "audio_tokens": null,
      "reasoning_tokens": 448
    },
    "prompt_tokens_details": {
      "audio_tokens": null,
      "cached_tokens": 0
    }
  },
  "prompt_filter_results": [
    {
      "prompt_index": 0,
      "content_filter_results": {
        "custom_blocklists": {
          "filtered": false
        },
        "hate": {
          "filtered": false,
          "severity": "safe"
        },
        "jailbreak": {
          "filtered": false,
          "detected": false
        },
        "self_harm": {
          "filtered": false,
          "severity": "safe"
        },
        "sexual": {
          "filtered": false,
          "severity": "safe"
        },
        "violence": {
          "filtered": false,
          "severity": "safe"
        }
      }
    }
  ]
}

推理工作

开发人员消息

从功能上来说，开发人员消息 "role": "developer" 与系统消息相同。

在上一代码示例中添加开发人员消息，如下所示：

pip install openai --upgrade

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.getenv("AZURE_OPENAI_API_KEY"),
    base_url="https://YOUR-RESOURCE-NAME.openai.azure.com/openai/v1/",
)

response = client.chat.completions.create(
    model="gpt-5-mini", # replace with the model deployment name of your o1 deployment.
    messages=[
        {"role": "developer","content": "You are a helpful assistant."}, # optional equivalent to a system message for reasoning models 
        {"role": "user", "content": "What steps should I think about when writing my first Python API?"},
    ],
    max_completion_tokens = 5000,
    reasoning_effort = "medium" # low, medium, or high
)

print(response.model_dump_json(indent=2))

pip install openai --upgrade

from openai import OpenAI
from azure.identity import DefaultAzureCredential, get_bearer_token_provider

token_provider = get_bearer_token_provider(
    DefaultAzureCredential(), "https://cognitiveservices.azure.com/.default"
)

client = OpenAI(  
  base_url = "https://YOUR-RESOURCE-NAME.openai.azure.com/openai/v1/",  
  api_key=token_provider,
)

response = client.chat.completions.create(
    model="o1-new", # replace with your model deployment name 
    messages=[
        {"role": "developer","content": "You are a helpful assistant."}, # optional equivalent to a system message for reasoning models 
        {"role": "user", "content": "What steps should I think about when writing my first Python API?"},
    ],
    max_completion_tokens = 5000,
    reasoning_effort = "medium" # low, medium, or high

)

print(response.model_dump_json(indent=2))

using Azure.Identity;
using OpenAI;
using OpenAI.Chat;
using System.ClientModel.Primitives;

#pragma warning disable OPENAI001 //currently required for token based authentication

BearerTokenPolicy tokenPolicy = new(
    new DefaultAzureCredential(),
    "https://cognitiveservices.azure.com/.default");

ChatClient client = new(
    model: "o4-mini",
    authenticationPolicy: tokenPolicy,
    options: new OpenAIClientOptions()
    {

        Endpoint = new Uri("https://YOUR-RESOURCE-NAME.openai.azure.com/openai/v1")
    }
);

ChatCompletionOptions options = new ChatCompletionOptions
{
    ReasoningEffortLevel = ChatReasoningEffortLevel.Low,
    MaxOutputTokenCount = 100000
};

ChatCompletion completion = client.CompleteChat(
         new DeveloperChatMessage("You are a helpful assistant"),
         new UserChatMessage("Tell me about the bitter lesson")
    );

Console.WriteLine($"[ASSISTANT]: {completion.Content[0].Text}");

推理摘要

将最新的推理模型与响应 API 结合使用时，可以使用推理摘要参数来接收模型思维链推理的摘要。

pip install openai --upgrade

import os
from openai import OpenAI

client = OpenAI(  
  base_url = "https://YOUR-RESOURCE-NAME.openai.azure.com/openai/v1/",
  api_key=os.getenv("AZURE_OPENAI_API_KEY")  
)

response = client.responses.create(
    input="Tell me about the curious case of neural text degeneration",
    model="gpt-5", # replace with model deployment name
    reasoning={
        "effort": "medium",
        "summary": "auto" # auto, concise, or detailed, gpt-5 series do not support concise 
    },
    text={
        "verbosity": "low" # New with GPT-5 models
    }
)

print(response.model_dump_json(indent=2))

curl -X POST "https://YOUR-RESOURCE-NAME.openai.azure.com/openai/v1/responses" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $AZURE_OPENAI_AUTH_TOKEN" \
 -d '{
     "model": "gpt-5",
     "input": "Tell me about the curious case of neural text degeneration",
     "reasoning": {"summary": "auto"},
     "text": {"verbosity": "low"}
    }'

{
  "id": "resp_689a0a3090808190b418acf12b5cc40e0fc1c31bc69d8719",
  "created_at": 1754925616.0,
  "error": null,
  "incomplete_details": null,
  "instructions": null,
  "metadata": {},
  "model": "gpt-5",
  "object": "response",
  "output": [
    {
      "id": "rs_689a0a329298819095d90c34dc9b80db0fc1c31bc69d8719",
      "summary": [],
      "type": "reasoning",
      "encrypted_content": null,
      "status": null
    },
    {
      "id": "msg_689a0a33009881909fe0fcf57cba30200fc1c31bc69d8719",
      "content": [
        {
          "annotations": [],
          "text": "Neural text degeneration refers to the ways language models produce low-quality, repetitive, or vacuous text, especially when generating long outputs. It’s “curious” because models trained to imitate fluent text can still spiral into unnatural patterns. Key aspects:\n\n- Repetition and loops: The model repeats phrases or sentences (“I’m sorry, but...”), often due to high-confidence tokens reinforcing themselves.\n- Loss of specificity: Vague, generic, agreeable text that avoids concrete details.\n- Drift and contradiction: The output gradually departs from context or contradicts itself over long spans.\n- Exposure bias: During training, models see gold-standard prefixes; at inference, they must condition on their own imperfect outputs, compounding errors.\n- Likelihood vs. quality mismatch: Maximizing token-level likelihood doesn’t align with human preferences for diversity, coherence, or factuality.\n- Token over-optimization: Frequent, safe tokens get overused; certain phrases become attractors.\n- Entropy collapse: With greedy or low-temperature decoding, the distribution narrows too much, causing repetitive, low-entropy text.\n- Length and beam search issues: Larger beams or long generations can favor bland, repetitive sequences (the “likelihood trap”).\n\nCommon mitigations:\n\n- Decoding strategies:\n  - Top-k, nucleus (top-p), or temperature sampling to keep sufficient entropy.\n  - Typical sampling and locally typical sampling to avoid dull but high-probability tokens.\n  - Repetition penalties, presence/frequency penalties, no-repeat n-grams.\n  - Contrastive decoding (and variants like DoLa) to filter generic continuations.\n  - Min/max length, stop sequences, and beam search with diversity/penalties.\n\n- Training and alignment:\n  - RLHF/DPO to better match human preferences for non-repetitive, helpful text.\n  - Supervised fine-tuning on high-quality, diverse data; instruction tuning.\n  - Debiasing objectives (unlikelihood training) to penalize repetition and banned patterns.\n  - Mixture-of-denoisers or latent planning to improve long-range coherence.\n\n- Architectural and planning aids:\n  - Retrieval-augmented generation to ground outputs.\n  - Tool use and structured prompting to constrain drift.\n  - Memory and planning modules, hierarchical decoding, or sentence-level control.\n\n- Prompting tips:\n  - Ask for concise answers, set token limits, and specify structure.\n  - Provide concrete constraints or content to reduce generic filler.\n  - Use “say nothing if uncertain” style instructions to avoid vacuity.\n\nRepresentative papers/terms to search:\n- Holtzman et al., “The Curious Case of Neural Text Degeneration” (2020): nucleus sampling.\n- Welleck et al., “Neural Text Degeneration with Unlikelihood Training.”\n- Li et al., “A Contrastive Framework for Decoding.”\n- Su et al., “DoLa: Decoding by Contrasting Layers.”\n- Meister et al., “Typical Decoding.”\n- Ouyang et al., “Training language models to follow instructions with human feedback.”\n\nIn short, degeneration arises from a mismatch between next-token likelihood and human preferences plus decoding choices; careful decoding, training objectives, and grounding help prevent it.",
          "type": "output_text",
          "logprobs": null
        }
      ],
      "role": "assistant",
      "status": "completed",
      "type": "message"
    }
  ],
  "parallel_tool_calls": true,
  "temperature": 1.0,
  "tool_choice": "auto",
  "tools": [],
  "top_p": 1.0,
  "background": false,
  "max_output_tokens": null,
  "max_tool_calls": null,
  "previous_response_id": null,
  "prompt": null,
  "prompt_cache_key": null,
  "reasoning": {
    "effort": "minimal",
    "generate_summary": null,
    "summary": "detailed"
  },
  "safety_identifier": null,
  "service_tier": "default",
  "status": "completed",
  "text": {
    "format": {
      "type": "text"
    }
  },
  "top_logprobs": null,
  "truncation": "disabled",
  "usage": {
    "input_tokens": 16,
    "input_tokens_details": {
      "cached_tokens": 0
    },
    "output_tokens": 657,
    "output_tokens_details": {
      "reasoning_tokens": 0
    },
    "total_tokens": 673
  },
  "user": null,
  "content_filters": null,
  "store": true
}

Python lark

GPT-5 系列推理模型能够调用名为 custom_tool 的新 lark_tool。 该工具基于 Python lark 开发，可用于对模型输出进行更灵活的限制。

响应 API

{
  "model": "gpt-5-2025-08-07",
  "input": "please calculate the area of a circle with radius equal to the number of 'r's in strawberry",
  "tools": [
    {
      "type": "custom",
      "name": "lark_tool",
      "format": {
        "type": "grammar",
        "syntax": "lark",
        "definition": "start: QUESTION NEWLINE ANSWER\nQUESTION: /[^\\n?]{1,200}\\?/\nNEWLINE: /\\n/\nANSWER: /[^\\n!]{1,200}!/"
      }
    }
  ],
  "tool_choice": "required"
}

import os
from openai import OpenAI

client = OpenAI(  
  base_url = "https://YOUR-RESOURCE-NAME.openai.azure.com/openai/v1/",
  api_key=os.getenv("AZURE_OPENAI_API_KEY")  
)

response = client.responses.create(  
    model="gpt-5",  # replace with your model deployment name  
    tools=[  
        {  
            "type": "custom",
            "name": "lark_tool",
            "format": {
                "type": "grammar",
                "syntax": "lark",
                "definition": "start: QUESTION NEWLINE ANSWER\nQUESTION: /[^\\n?]{1,200}\\?/\nNEWLINE: /\\n/\nANSWER: /[^\\n!]{1,200}!/"
            }
        }  
    ],  
    input=[{"role": "user", "content": "Please calculate the area of a circle with radius equal to the number of 'r's in strawberry"}],  
)  

print(response.model_dump_json(indent=2))  
  

输出：

{
  "id": "resp_689a0cf927408190b8875915747667ad01c936c6ffb9d0d3",
  "created_at": 1754926332.0,
  "error": null,
  "incomplete_details": null,
  "instructions": null,
  "metadata": {},
  "model": "gpt-5",
  "object": "response",
  "output": [
    {
      "id": "rs_689a0cfd1c888190a2a67057f471b5cc01c936c6ffb9d0d3",
      "summary": [],
      "type": "reasoning",
      "encrypted_content": null,
      "status": null
    },
    {
      "id": "msg_689a0d00e60c81908964e5e9b2d6eeb501c936c6ffb9d0d3",
      "content": [
        {
          "annotations": [],
          "text": "“strawberry” has 3 r’s, so the radius is 3.\nArea = πr² = π × 3² = 9π ≈ 28.27 square units.",
          "type": "output_text",
          "logprobs": null
        }
      ],
      "role": "assistant",
      "status": "completed",
      "type": "message"
    }
  ],
  "parallel_tool_calls": true,
  "temperature": 1.0,
  "tool_choice": "auto",
  "tools": [
    {
      "name": "lark_tool",
      "parameters": null,
      "strict": null,
      "type": "custom",
      "description": null,
      "format": {
        "type": "grammar",
        "definition": "start: QUESTION NEWLINE ANSWER\nQUESTION: /[^\\n?]{1,200}\\?/\nNEWLINE: /\\n/\nANSWER: /[^\\n!]{1,200}!/",
        "syntax": "lark"
      }
    }
  ],
  "top_p": 1.0,
  "background": false,
  "max_output_tokens": null,
  "max_tool_calls": null,
  "previous_response_id": null,
  "prompt": null,
  "prompt_cache_key": null,
  "reasoning": {
    "effort": "medium",
    "generate_summary": null,
    "summary": null
  },
  "safety_identifier": null,
  "service_tier": "default",
  "status": "completed",
  "text": {
    "format": {
      "type": "text"
    }
  },
  "top_logprobs": null,
  "truncation": "disabled",
  "usage": {
    "input_tokens": 139,
    "input_tokens_details": {
      "cached_tokens": 0
    },
    "output_tokens": 240,
    "output_tokens_details": {
      "reasoning_tokens": 192
    },
    "total_tokens": 379
  },
  "user": null,
  "content_filters": null,
  "store": true
}

聊天完成

{
  "messages": [
    {
      "role": "user",
      "content": "Which one is larger, 42 or 0?"
    }
  ],
  "tools": [
    {
      "type": "custom",
      "name": "custom_tool",
      "custom": {
        "name": "lark_tool",
        "format": {
          "type": "grammar",
          "grammar": {
            "syntax": "lark",
            "definition": "start: QUESTION NEWLINE ANSWER\nQUESTION: /[^\\n?]{1,200}\\?/\nNEWLINE: /\\n/\nANSWER: /[^\\n!]{1,200}!/"
          }
        }
      }
    }
  ],
  "tool_choice": "required",
  "model": "gpt-5-2025-08-07"
}

Markdown 输出

默认情况下，o3-mini 和 o1 模型不会尝试生成包含 markdown 格式的输出。 一个常见的使用场景是，当希望模型输出包含在 markdown 代码块中的代码时，这种行为是不理想的。 当模型生成不带 markdown 格式的输出时，会在交互式操场体验中丢失语法突出显示和可复制代码块等功能。 若要替代此新的默认行为并鼓励在模型响应中包含 Markdown，请将字符串 Formatting re-enabled 添加到开发人员消息的开头。

将 Formatting re-enabled 添加到开发人员消息的开头不能保证模型在其响应中包含 markdown 格式，只会增加其可能性。 我们在内部测试中发现，具有 Formatting re-enabled 模型的 o1 与具有 o3-mini 相比，其本身效率更低。

为了提高 Formatting re-enabled 的性能，可以进一步增强开发人员信息的开头部分，这通常会产生所需的输出结果。 可以尝试添加更具描述性的初始说明，而不是将 Formatting re-enabled 添加到开发人员消息的开头，如以下示例之一：

• Formatting re-enabled - please enclose code blocks with appropriate markdown tags.
• Formatting re-enabled - code output should be wrapped in markdown.

根据预期输出，可能需要进一步自定义初始开发人员消息，以针对特定用例。