便携AI今天OpenAI推出了全新的o3-pro模型，这是继o3和o3‑mini之后，专为追求“可靠性”而非速度的大型推理模型，通过深度思考，利用更大算力和更长思考时间，o3‑pro有能力处理复杂逻辑推理，比如数学、科学、编程、商业决策等，目前便携AI聚合API已经支持最新的o3-pro模型，本文分享下调用方法。
一、o3-pro模型介绍
o3-pro可以进行“更深的思考”，基于o3架构，增加推理计算，优化链式思考，提升复杂任务表现，官方建议将o3‑pro用于“高挑战、高价值、需要高度准确性且时间延迟可以接受”的任务。
以下是OpenAI官方对于o3-pro的介绍：
The o-series of models are trained with reinforcement learning to think before they answer and perform complex reasoning. The o3-pro model uses more compute to think harder and provide consistently better answers.
o3-pro is available in the Responses API only to enable support for multi-turn model interactions before responding to API requests, and other advanced API features in the future. Since o3-pro is designed to tackle tough problems, some requests may take several minutes to finish. To avoid timeouts, try using background mode.
翻译：
o系列模型通过强化学习进行训练，具备“先思考再作答”的能力，能够执行复杂的推理任务。
其中，o3-pro模型使用了更多计算资源，以便“更深入地思考”，从而提供更稳定、更优质的回答。o3-pro 仅在 Responses API 中提供，旨在支持模型在响应 API 请求前进行多轮对话交互，并为未来的高级 API 功能奠定基础。
由于 o3-pro 专为解决复杂问题而设计，部分请求可能需要几分钟才能完成。为了避免请求超时，建议使用后台模式（background mode）。
总的来说，o3-pro思考更久，能力更强，可以进行更复杂的推理任务，但是非常耗时。
o3-pro价格如下：
- 输入token：$20/1M tokens
- 输出token：$80/1M tokens
二、o3-pro模型调用方法
o3-pro模型名称：o3-pro、o3-pro-2025-06-10
o3-pro只能使用Responses API，传统的Chat Completions API无法调用。
下面以Python为例，介绍下如何调用o3-pro。示例中的api_key可以在网站后台获取，获取方法：《便携AI聚合API新建令牌（API key）教程》。
1、OpenAI包调用
def response_openai():
from openai import OpenAI
client = OpenAI(
api_key="sk-234345",
base_url=f'https://api.bianxie.ai/v1'
)
response = client.responses.create(
model="o3-pro",
input=[
{
"role": "user",
"content": "Write a one-sentence bedtime story about a unicorn."
}
]
)
print(response.json())
返回示例：
{
"id": "resp_6849735d219881a28f8a567b1d3070550949fa752fa96f08",
"created_at": 1749644125,
"error": null,
"incomplete_details": null,
"instructions": null,
"metadata": {
},
"model": "o3-pro-2025-06-10",
"object": "response",
"output": [
{
"id": "rs_68497370d9f881a2a4c639f73758a5cb0949fa752fa96f08",
"summary": [
],
"type": "reasoning",
"status": null
},
{
"id": "msg_68497370daf481a29dc7ba6a5fd71b690949fa752fa96f08",
"content": [
{
"annotations": [
],
"text": "As moonlight painted the clouds silver, a gentle unicorn tiptoed through dream-dust meadows, collecting children’s worries in her shimmering mane and replacing them with whispers of starlit peace.",
"type": "output_text"
}
],
"role": "assistant",
"status": "in_progress",
"type": "message"
}
],
"parallel_tool_calls": true,
"temperature": 1,
"tool_choice": "auto",
"tools": [
],
"top_p": 1,
"max_output_tokens": null,
"previous_response_id": null,
"reasoning": {
"effort": "medium",
"generate_summary": null,
"summary": null
},
"service_tier": "default",
"status": "completed",
"text": {
"format": {
"type": "text"
}
},
"truncation": "disabled",
"usage": {
"input_tokens": 17,
"input_tokens_details": {
"cached_tokens": 0
},
"output_tokens": 46,
"output_tokens_details": {
"reasoning_tokens": 0
},
"total_tokens": 63
},
"user": null,
"background": false,
"store": true
}
2、curl调用
import requests
url = f"https://api.bianxie.ai/v1/responses"
headers = {
"Content-Type": "application/json",
"Authorization": f"Bearer sk-2jsodfjo"
}
data = {
"model": "o3-pro",
"input": "Tell me a three sentence bedtime story about a unicorn."
}
response = requests.post(url, headers=headers, json=data)
print(response.json())
返回示例：
{
"id": "resp_684977361f8081a18b547b259f2d72c20ffee4a22ce28380",
"object": "response",
"created_at": 1749645110,
"status": "completed",
"background": false,
"error": null,
"incomplete_details": null,
"instructions": null,
"max_output_tokens": null,
"model": "o3-pro-2025-06-10",
"output": [
{
"id": "rs_6849774897a081a1ac739a829fd56a230ffee4a22ce28380",
"type": "reasoning",
"summary": []
},
{
"id": "msg_68497748998881a1b8868097f2ed564e0ffee4a22ce28380",
"type": "message",
"status": "in_progress",
"content": [
{
"type": "output_text",
"annotations": [],
"text": "As Luna the silver-maned unicorn tiptoed through the moonlit forest, her horn softly glowed and stitched tiny stars back into holes in the night sky. Each twinkling stitch hummed a lullaby that drifted into every woodland burrow, lulling even the busiest fireflies to sleep. When her work was done, Luna curled beside a quiet pond, and the sky, now perfectly mended, shimmered like a gentle blanket over all the dreaming creatures below."
}
],
"role": "assistant"
}
],
"parallel_tool_calls": true,
"previous_response_id": null,
"reasoning": {
"effort": "medium",
"summary": null
},
"service_tier": "default",
"store": true,
"temperature": 1.0,
"text": {
"format": {
"type": "text"
}
},
"tool_choice": "auto",
"tools": [],
"top_p": 1.0,
"truncation": "disabled",
"usage": {
"input_tokens": 17,
"input_tokens_details": {
"cached_tokens": 0
},
"output_tokens": 104,
"output_tokens_details": {
"reasoning_tokens": 0
},
"total_tokens": 121
},
"user": null,
"metadata": {}
}
