GPT-5 Pro API: Model ID, Setup, Pricing, and Deprecation

Yes—OpenAI documents GPT-5 Pro for API use. Its model ID is gpt-5-pro, and it is supported through the Responses API, not Chat Completions. However, OpenAI marks GPT-5 Pro as deprecated, so it may be a poor choice for a new production integration. If you already have access and need to test an existing system, the example below shows a basic request. (OpenAI GPT-5 Pro model documentation)

GPT-5 Pro API details at a glance

Item	OpenAI documentation
Model ID	gpt-5-pro
Status	Deprecated
Supported API	Responses API (/v1/responses) only
Reasoning effort	high only
Input and output	Text and images in; text out
Context window	400,000 tokens
Maximum output	272,000 tokens
Price	$15 per 1 million input tokens; $120 per 1 million output tokens
Latency note	Some requests may take several minutes

These are GPT-5 Pro-specific details from OpenAI’s model page. The model page also says GPT-5 Pro does not support Code Interpreter. Rate limits depend on usage tier, so check the limits for the API project that will make the calls. (OpenAI GPT-5 Pro model documentation)

Make a first GPT-5 Pro API request

You need an OpenAI API key and the OpenAI Python SDK. OpenAI’s quickstart shows creating a key, storing it in an environment variable, installing the SDK, and calling client.responses.create. Keep the key on your server or in your local environment—don’t put it in browser code or commit it to source control. (OpenAI API quickstart)

1. Install the SDK:

   pip install openai
   

2. Set your API key in the terminal. On macOS or Linux:

   export OPENAI_API_KEY="your_api_key_here"
   

   In PowerShell:

   $env:OPENAI_API_KEY="your_api_key_here"
   

3. Save this as example.py and run python example.py:

   from openai import OpenAI
   
   client = OpenAI()
   
   response = client.responses.create(
       model="gpt-5-pro",
       input="Give me three concise checks for reviewing a database migration plan.",
   )
   
   print(response.output_text)
   

The request uses the documented model ID and Responses API. A successful call returns a response whose text you can read with response.output_text. Because OpenAI marks this model deprecated, treat the snippet as a basic integration test—not a recommendation to build a new production system on it. Your project must also have access to the model for the request to succeed. (OpenAI GPT-5 Pro model documentation)

Why use Responses rather than Chat Completions?

GPT-5 Pro is listed as supported on the Responses API and not supported on Chat Completions. Don’t substitute a Chat Completions endpoint or assume that an endpoint supported by a different GPT model will work for Pro. OpenAI recommends the Responses API for new integrations generally. (OpenAI GPT-5 Pro model documentation)

Pricing and production fit

OpenAI lists GPT-5 Pro at $15 per 1 million input tokens and $120 per 1 million output tokens. For a simple estimate, a request with 10,000 billed input tokens and 2,000 billed output tokens would cost $0.39 at those rates: (10,000 ÷ 1,000,000 × $15) + (2,000 ÷ 1,000,000 × $120). This is an arithmetic example, not a measured request cost; actual usage depends on billed tokens, and separate tool charges may apply when using chargeable tools. (OpenAI GPT-5 Pro model documentation)

GPT-5 Pro is intended for demanding reasoning tasks, but its documented constraints matter: it supports only high reasoning effort, and some requests may take several minutes. The model page lists image and text input, text output, and no Code Interpreter support. Test representative tasks for quality, latency, and spend before relying on it in production; a high-level capability description alone does not establish performance for your workload. (OpenAI GPT-5 Pro model documentation)

Should you use GPT-5 Pro for a new integration?

Usually, start by reviewing OpenAI’s current model guidance rather than choosing GPT-5 Pro by name. OpenAI’s guidance describes GPT-6 Astra for the most demanding reasoning and professional work, and GPT-6.1 Sol as a lower-cost option for complex work. Compare currently supported models against your own quality, latency, and cost requirements; don’t assume GPT-5 Pro’s model ID, price, or limits apply to another model. (OpenAI current model guidance)

Bottom line: GPT-5 Pro has a documented API model ID (gpt-5-pro) and a supported Responses API path, but OpenAI marks it deprecated. Use it only if it remains available to your project and you have a specific reason to maintain or test it. For a new application, evaluate a currently recommended model instead. (OpenAI GPT-5 Pro model documentation)

Sources

GPT-5 Pro Model | OpenAI API

Developer quickstart | OpenAI API

Migrate to the Responses API | OpenAI API

Using GPT-6 | OpenAI API