OpenAI Chat Completion API Application and Usage

OpenAI ChatGPT is a very powerful AI dialogue system that can generate smooth and natural replies in just a few seconds by inputting prompts. ChatGPT stands out in the industry with its excellent language understanding and generation capabilities, and today, it has been widely applied across various industries and fields, with its influence becoming increasingly significant. Whether for daily conversations, creative writing, or professional consulting and coding, ChatGPT can provide astonishing intelligent assistance, greatly enhancing human work efficiency and creativity.

This document mainly introduces the usage process of the OpenAI Chat Completion API, allowing us to easily utilize the dialogue features of the official OpenAI ChatGPT.

Application Process

To use the OpenAI Chat Completion API, first go to the Ace Data Cloud Console to obtain your API Token for backup.

If you are not logged in or registered, you will be automatically redirected to the login page to invite you to register and log in, and after completion, you will be automatically returned to the current page.

One API Token can call all services on the platform without needing to apply separately for each service. The first application will grant a free quota for a free experience; when the quota is insufficient, you can recharge the general balance in the console.

📘 Complete Documentation: OpenAI Chat Completion API →

Basic Usage

Next, you can fill in the corresponding content on the interface, as shown in the figure:

When using this interface for the first time, we need to fill in at least three pieces of content: one is authorization, which can be selected directly from the dropdown list. The other parameter is model, which is the category of the OpenAI ChatGPT model we choose to use; here we mainly have 20 types of models, and details can be found in the models we provide. The last parameter is messages, which is an array of our input questions; it is an array that allows multiple questions to be uploaded simultaneously, with each question containing role and content, where role indicates the role of the questioner, and we provide three identities: user, assistant, and system. The other content is the specific content of our question.

You can also notice that there is corresponding code generation on the right side; you can copy the code to run directly or click the "Try" button for testing.

Common optional parameters:

  • max_tokens: Limits the maximum number of tokens for a single reply.
  • temperature: Generates randomness, between 0-2, with larger values being more divergent.
  • n: How many candidate replies to generate at once.
  • response_format: Sets the return format.

After the call, we find the return result as follows:

{
  "id": "chatcmpl-Cmd6uwSxN75F4PAdQSFEO8f2QPs4E",
  "object": "chat.completion",
  "created": 1765706120,
  "model": "gpt-5.5",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Hello! What can I help you with today?",
        "refusal": null,
        "annotations": []
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 7,
    "completion_tokens": 13,
    "total_tokens": 20,
    "prompt_tokens_details": {
      "cached_tokens": 0,
      "audio_tokens": 0
    },
    "completion_tokens_details": {
      "reasoning_tokens": 0,
      "audio_tokens": 0,
      "accepted_prediction_tokens": 0,
      "rejected_prediction_tokens": 0
    }
  },
  "service_tier": "default",
  "system_fingerprint": null
}

The return result contains multiple fields, described as follows:

  • id: The ID generated for this dialogue task, used to uniquely identify this dialogue task.
  • model: The selected OpenAI ChatGPT model.
  • choices: The response information provided by ChatGPT for the question.
  • usage: Statistics on token usage for this Q&A.

Among them, choices contains the response information from ChatGPT, and within it, the choices is ChatGPT, as shown in the figure.

As can be seen, the content field in choices contains the specific content of ChatGPT's reply.

Streaming Response

This interface also supports streaming responses, which is very useful for web integration, allowing the webpage to achieve a word-by-word display effect.

If you want to return responses in a streaming manner, you can change the stream parameter in the request header to true.

Modify as shown in the figure, but the calling code needs to have corresponding changes to support streaming responses.

After changing stream to true, the API will return the corresponding JSON data line by line, and we need to make corresponding modifications at the code level to obtain the line-by-line results.

Python sample calling code:

import requests

url = "https://api.acedata.cloud/openai/chat/completions"

headers = {
    "accept": "application/json",
    "authorization": "Bearer {token}",
    "content-type": "application/json"
}

payload = {
    "model": "gpt-4",
    "messages": [{"role":"user","content":"hello"}],
    "stream": True
}

response = requests.post(url, json=payload, headers=headers)
print(response.text)

The output effect is as follows:

data: {"choices": [{"delta": {"role": "assistant"}, "index": 0}], "created": 1721007348, "id": "chatcmpl-YzczYjVhNjhjMzMwNDQ5MDkyNGYzOGZjZGE1ZGQ5OGU", "model": "gpt-4", "object": "chat.completion.chunk", "recipient": "all"}

data: {"choices": [{"delta": {"content": "Hi", "role": "assistant"}, "index": 0}], "created": 1721007348, "id": "chatcmpl-YzczYjVhNjhjMzMwNDQ5MDkyNGYzOGZjZGE1ZGQ5OGU", "model": "gpt-4", "object": "chat.completion.chunk", "recipient": "all"}

data: {"choices": [{"delta": {"content": " there", "role": "assistant"}, "index": 0}], "created": 1721007348, "id": "chatcmpl-YzczYjVhNjhjMzMwNDQ5MDkyNGYzOGZjZGE1ZGQ5OGU", "model": "gpt-4", "object": "chat.completion.chunk", "recipient": "all"}

data: {"choices": [{"delta": {"content": "!", "role": "assistant"}, "index": 0}], "created": 1721007348, "id": "chatcmpl-YzczYjVhNjhjMzMwNDQ5MDkyNGYzOGZjZGE1ZGQ5OGU", "model": "gpt-4", "object": "chat.completion.chunk", "recipient": "all"}

data: {"choices": [{"delta": {"content": " How", "role": "assistant"}, "index": 0}], "created": 1721007348, "id": "chatcmpl-YzczYjVhNjhjMzMwNDQ5MDkyNGYzOGZjZGE1ZGQ5OGU", "model": "gpt-4", "object": "chat.completion.chunk", "recipient": "all"}

data: {"choices": [{"delta": {"content": " can", "role": "assistant"}, "index": 0}], "created": 1721007348, "id": "chatcmpl-YzczYjVhNjhjMzMwNDQ5MDkyNGYzOGZjZGE1ZGQ5OGU", "model": "gpt-4", "object": "chat.completion.chunk", "recipient": "all"}

data: {"choices": [{"delta": {"content": " I", "role": "assistant"}, "index": 0}], "created": 1721007348, "id": "chatcmpl-YzczYjVhNjhjMzMwNDQ5MDkyNGYzOGZjZGE1ZGQ5OGU", "model": "gpt-4", "object": "chat.completion.chunk", "recipient": "all"}

data: {"choices": [{"delta": {"content": " assist", "role": "assistant"}, "index": 0}], "created": 1721007348, "id": "chatcmpl-YzczYjVhNjhjMzMwNDQ5MDkyNGYzOGZjZGE1ZGQ5OGU", "model": "gpt-4", "object": "chat.completion.chunk", "recipient": "all"}

data: {"choices": [{"delta": {"content": " you", "role": "assistant"}, "index": 0}], "created": 1721007348, "id": "chatcmpl-YzczYjVhNjhjMzMwNDQ5MDkyNGYzOGZjZGE1ZGQ5OGU", "model": "gpt-4", "object": "chat.completion.chunk", "recipient": "all"}

data: {"choices": [{"delta": {"content": " today", "role": "assistant"}, "index": 0}], "created": 1721007348, "id": "chatcmpl-YzczYjVhNjhjMzMwNDQ5MDkyNGYzOGZjZGE1ZGQ5OGU", "model": "gpt-4", "object": "chat.completion.chunk", "recipient": "all"}

data: {"choices": [{"delta": {"content": "?", "role": "assistant"}, "index": 0}], "created": 1721007348, "id": "chatcmpl-YzczYjVhNjhjMzMwNDQ5MDkyNGYzOGZjZGE1ZGQ5OGU", "model": "gpt-4", "object": "chat.completion.chunk", "recipient": "all"}

data: {"choices": [{"delta": {"role": "assistant"}, "index": 0}], "created": 1721007348, "id": "chatcmpl-YzczYjVhNjhjMzMwNDQ5MDkyNGYzOGZjZGE1ZGQ5OGU", "model": "gpt-4", "object": "chat.completion.chunk", "recipient": "all"}

data: {"choices": [{"delta": {"role": "assistant"}, "finish_reason": "stop", "index": 0}], "created": 1721007349, "id": "chatcmpl-YzczYjVhNjhjMzMwNDQ5MDkyNGYzOGZjZGE1ZGQ5OGU", "model": "gpt-4", "object": "chat.completion.chunk", "recipient": "all"}

data: [DONE]

It can be seen that there are many data in the response, and the choices in data are the latest response content, consistent with the content introduced above. The choices are the newly added response content, which you can use to connect to your system. At the same time, the end of the streaming response is determined by the content of data. If the content is [DONE], it indicates that the streaming response has completely ended. The returned data result has multiple fields, which are described as follows:

  • id, the ID generated for this dialogue task, used to uniquely identify this dialogue task.
  • model, the OpenAI ChatGPT model selected.
  • choices, the response information provided by ChatGPT to the prompt.

JavaScript is also supported, for example, the streaming call code for Node.js is as follows:

const options = {
  method: "post",
  headers: {
    accept: "application/json",
    authorization: "Bearer {token}",
    "content-type": "application/json",
  },
  body: JSON.stringify({
    model: "gpt-4",
    messages: [{ role: "user", content: "hello" }],
    stream: true,
  }),
};

fetch("https://api.acedata.cloud/openai/chat/completions", options)
  .then((response) => response.json())
  .then((response) => console.log(response))
  .catch((err) => console.error(err));

Java sample code:

JSONObject jsonObject = new JSONObject();
jsonObject.put("model", "gpt-4");
jsonObject.put("messages", [{"role":"user","content":"hello"}]);
jsonObject.put("stream", true);
MediaType mediaType = "application/json; charset=utf-8".toMediaType();
RequestBody body = jsonObject.toString().toRequestBody(mediaType);
Request request = new Request.Builder()
  .url("https://api.acedata.cloud/openai/chat/completions")
  .post(body)
  .addHeader("accept", "application/json")
  .addHeader("authorization", "Bearer {token}")
  .addHeader("content-type", "application/json")
  .build();

OkHttpClient client = new OkHttpClient();
Response response = client.newCall(request).execute();
System.out.print(response.body!!.string())

Other languages can be rewritten accordingly; the principle is the same.

Multi-turn Dialogue

If you want to connect to the multi-turn dialogue feature, you need to upload multiple prompts in the messages field. The specific examples of multiple prompts are shown in the image below:

Python sample call code:

import requests

url = "https://api.acedata.cloud/openai/chat/completions"

headers = {
    "accept": "application/json",
    "authorization": "Bearer {token}",
    "content-type": "application/json"
}

payload = {
    "model": "gpt-4",
    "messages": [{"role":"user","content":"Hello"},{"role":"assistant","content":"Hi! How can I assist you today?"},{"role":"user","content":"What I say just now?"}]
}

response = requests.post(url, json=payload, headers=headers)
print(response.text)

By uploading multiple question words, multi-turn dialogue can be easily achieved, resulting in the following response:

{
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "You said, \"Hello.\""
      },
      "finish_reason": "stop"
    }
  ],
  "created": 1721323012,
  "id": "chatcmpl-NWZmOTA5MDlkZjBjNDRjNGEwMzRjYzA5NmM1MzQwMWY",
  "model": "gpt-4",
  "object": "chat.completion.chunk",
  "recipient": "all",
  "usage": {
    "prompt_tokens": 31,
    "completion_tokens": 6,
    "total_tokens": 37
  }
}

As can be seen, the information contained in choices is consistent with the basic usage content, which includes the specific content of ChatGPT's responses to multiple dialogues, allowing for answers to corresponding questions based on multiple dialogue contents.

Integrating OpenAI-Python

The OpenAI Chat Completion API is compatible with the official OpenAI interface and can be directly integrated using the official SDK OpenAI-Python. This article will briefly introduce the usage.

  1. First, set up a local Python environment, which can be searched on Google.
  2. Download and install a development environment, such as the VSCode editor.
  3. Configure the OpenAI environment variables.
  • In the project folder, create a file named .env and save it.
  • The content of the .env file:
OPENAI_API_KEY="sk-xxx"
OPENAI_BASE_URL="https://api.acedata.cloud/openai"  # Reminder: If you are using the official OpenAI key, do not use this address.

Replace sk-xxx with your own key. OPENAI_BASE_URL is the proxy interface for accessing OpenAI.

  1. Install the project's dependency packages.
pip install openai

The command for Mac OS is:

pip3 install openai
  1. Create a sample source code file.

Assuming we create a sample code index.py, the specific content is as follows:

import os
from openai import OpenAI

client = OpenAI(api_key=os.environ.get("OPENAI_API_KEY"))

response = client.chat.completions.create(
    messages=[
        {
            "role": "user",
            "content": "hello",
        }
    ],
    model="gpt-4",
)

print(response.text)

Online Model

The gpt-3.5-browsing and gpt-4-browsing models are different from other models; they can perform online searches based on the question words and return the results of the online search with appropriate adjustments. This article will demonstrate the online functionality through a specific example, and you can fill in the corresponding content on the OpenAI Chat Completion API interface, as shown in the figure:

You can also notice that there is corresponding code generation on the right side; you can copy the code to run directly or click the "Try" button for testing.

After the call, we find that the returned result is as follows:

{
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "For the latest news in China today, you can check major news websites such as:\n\n- [BBC News China](https://www.bbc.com/news/world/asia/china)\n- [CNN China News](https://edition.cnn.com/china)\n- [Reuters China](https://www.reuters.com/news/archive/china-news)\n\nThese sources will have up-to-date information on current events in China."
      },
      "finish_reason": "stop"
    }
  ],
  "created": 1721009347,
  "id": "chatcmpl-YzA0M2RjZDVkYThlNDkxNTkzOThmZWQ4OGMzNzdhNzA",
  "model": "gpt-4-browsing",
  "object": "chat.completion.chunk",
  "recipient": "all",
  "usage": {
    "prompt_tokens": 325,
    "completion_tokens": 82,
    "total_tokens": 407
  }
}

As can be seen, the response information in choices is obtained based on online queries and also provides relevant links. The response information in choices needs to be rendered using markdown syntax to achieve the best experience, which ultimately reflects the powerful advantages of our model's online functionality.

Visual Model

gpt-4o is a multimodal large language model developed by OpenAI, which adds visual understanding capabilities on the basis of GPT-4. This model can process both text and image inputs simultaneously, achieving cross-modal understanding and generation.

The text processing using the gpt-4o model is consistent with the basic usage content mentioned above. Below, we will briefly introduce how to use the model's image processing capabilities.

The image processing capability of the gpt-4o model is mainly achieved by adding a type field to the original content, which indicates whether the uploaded content is text or an image, thus utilizing the image processing capabilities of the gpt-4o model. Below, we will mainly discuss how to call this function using both Curl and Python.

  • Curl script method
curl -X POST 'https://api.acedata.cloud/openai/chat/completions' \
-H 'accept: application/json' \
-H 'authorization: Bearer {token}' \
-H 'content-type: application/json' \
-d '{
    "model": "gpt-4o",
    "messages": [
      {
        "role": "user",
        "content": [
          {
            "type": "text",
            "text": "What'\''s in this image?"
          },
          {
            "type": "image_url",
            "image_url": {
              "url": "https://upload.wikimedia.org/wikipedia/commons/thumb/d/dd/Gfp-wisconsin-madison-the-nature-boardwalk.jpg/2560px-Gfp-wisconsin-madison-the-nature-boardwalk.jpg"
            }
          }
        ]
      }
    ]
  }'
  • Python script method
import requests

url = "https://api.acedata.cloud/openai/chat/completions"

headers = {
    "accept": "application/json",
    "authorization": "Bearer {token}",
    "content-type": "application/json"
}

payload = {
    "model": "gpt-4o",
    "messages": [
        {
            "role": "user",
            "content": [
                {
                    "type": "text", "text": "这张图片里有什么？"
                },
                {
                    "type": "image_url",
                    "image_url": {
                        "url": "https://upload.wikimedia.org/wikipedia/commons/thumb/d/dd/Gfp-wisconsin-madison-the-nature-boardwalk.jpg/2560px-Gfp-wisconsin-madison-the-nature-boardwalk.jpg"
                    }
                },
            ],
        }
    ]
}

response = requests.post(url, json=payload, headers=headers)
print(response.text)

Then you can get the following result, where the field information is consistent with the above text, specifically as follows:

{
  "id": "chatcmpl-123",
  "object": "chat.completion",
  "created": 1677652288,
  "model": "gpt-4-vision-preview",
  "system_fingerprint": "fp_44709d6fcb",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "\n\n这张图片展示了一条木栈道延伸穿过郁郁葱葱的沼泽地。"
      },
      "logprobs": null,
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 9,
    "completion_tokens": 12,
    "total_tokens": 21
  }
}

It can be seen that the content of the answer is based on the image, so through the above two methods, the text and image processing capabilities of the gpt-4-vision model can be easily utilized.

In addition to gpt-4o, there is a lower-cost model called gpt-4o-mini. gpt-4o-mini is the latest generation of large language models developed by OpenAI, which not only responds quickly but is also cheaper and supports multimodal capabilities. The use of vision features can refer to the content of the gpt-4o model mentioned above.

GPT-4o Drawing Model

Generate Images Based on Reference Images

Below is an example of generating a custom style image based on a picture. First, let’s take a look at the input image, as shown below:

It can be seen that the reference image is a real person's picture. We can ask it to change to a certain style, for example, to turn it into an anime-style image, with the specific request example:

{
  "model": "gpt-4o-image",
  "messages": [
    {
      "role": "user",
      "content": [
        {
          "type": "text",
          "text": "生成动漫风格的图片，并且带上个帽子"
        },
        {
          "type": "image_url",
          "image_url": {
            "url": "https://cdn.acedata.cloud/qzx2z1.png"
          }
        }
      ]
    }
  ],
  "stream": false
}

Sample result:

{
  "id": "chatcmpl-89DPQxbLuyRNzH5YLCPYM5WElV3dm",
  "object": "chat.completion",
  "created": 1781020664,
  "model": "gpt-4o-image",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "\n\n> 🎨 生成中...\n\n![https://platform.cdn.acedata.cloud/20260609/0f7b6cf1b14843b1bab8e261fe5765b3.png](https://platform.cdn.acedata.cloud/20260609/0f7b6cf1b14843b1bab8e261fe5765b3.png)\n\n[点击下载](https://platform.cdn.acedata.cloud/download/20260609/0f7b6cf1b14843b1bab8e261fe5765b3.png)"
      },
      "logprobs": null,
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 100,
    "completion_tokens": 122,
    "total_tokens": 222,
    "prompt_tokens_details": {
      "text_tokens": 93,
      "cached_tokens_details": {}
    },
    "completion_tokens_details": {}
  }
}

Among them, the choices in the message.content is the complete dialogue result generated, and the image is included in Markdown format (the image link is a temporary address, please download and save it in time). It can be seen that the generated image is indeed in anime style, as shown in the following image:

Pure Text Image Generation

We can generate an image through a prompt and return it to us in a conversational result. Below, we take create an image of a futuristic city at sunset as an example, with the specific example as follows:

{
  "model": "gpt-4o-image",
  "messages": [
    {
      "role": "user",
      "content": [
        {
          "type": "text",
          "text": "创建一张未来城市日落的图片"
        }
      ]
    }
  ],
  "stream": false
}

Sample result:

{
  "id": "chatcmpl-89DqkpQoPGkQqJ6kPKMKWejjLXVxQ",
  "object": "chat.completion",
  "created": 1781020587,
  "model": "gpt-4o-image",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "\n\n> 🎨 生成中...\n\n![https://platform.cdn.acedata.cloud/20260609/ed2cca68732540fc99162ddc10ddc153.png](https://platform.cdn.acedata.cloud/20260609/ed2cca68732540fc99162ddc10ddc153.png)\n\n[点击下载](https://platform.cdn.acedata.cloud/download/20260609/ed2cca68732540fc99162ddc10ddc153.png)"
      },
      "logprobs": null,
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 17,
    "completion_tokens": 104,
    "total_tokens": 121,
    "prompt_tokens_details": {
      "text_tokens": 10,
      "cached_tokens_details": {}
    },
    "completion_tokens_details": {}
  }
}

It can be seen that the result is consistent with the prompt, as shown below:

Generate One Image from Multiple Images

We can also use multiple reference images to generate one image. For example, using a handsome guy and a coffee image, these two images can be used to generate an image of a handsome guy drinking coffee. Below are the specific reference images:

Next, we take generate an image of a boy holding coffee and about to drink as an example, with the specific example as follows:

{
  "model": "gpt-4o-image",
  "messages": [
    {
      "role": "user",
      "content": [
        {
          "type": "text",
          "text": "Generate an image of a boy holding coffee and about to drink it."
        },
        {
          "type": "image_url",
          "image_url": {
            "url": "https://cdn.acedata.cloud/pqquv3.jpg"
          }
        },
        {
          "type": "image_url",
          "image_url": {
            "url": "https://cdn.acedata.cloud/h8j2i0.jpg"
          }
        }
      ]
    }
  ],
  "stream": false
}

Sample result:

{
  "id": "chatcmpl-89DnHbbzOIQvU1VzJrNjzMU8BRUgG",
  "object": "chat.completion",
  "created": 1781021018,
  "model": "gpt-4o-image",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "\n\n> 🎨 Generating...\n\n![https://platform.cdn.acedata.cloud/20260610/f1d9ddee3c304230a9f92929f04b95be.png](https://platform.cdn.acedata.cloud/20260610/f1d9ddee3c304230a9f92929f04b95be.png)\n\n[Click to download](https://platform.cdn.acedata.cloud/download/20260610/f1d9ddee3c304230a9f92929f04b95be.png)"
      },
      "logprobs": null,
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 193,
    "completion_tokens": 116,
    "total_tokens": 309,
    "prompt_tokens_details": {
      "text_tokens": 186,
      "cached_tokens_details": {}
    },
    "completion_tokens_details": {}
  }
}

As you can see, the generated result indeed combines the two images to create the output, here is the specific result:

Error Handling

When calling the API, if an error occurs, the API will return the corresponding error code and message. For example:

  • 400 token_mismatched: Bad request, possibly due to missing or invalid parameters.
  • 400 api_not_implemented: Bad request, possibly due to missing or invalid parameters.
  • 401 invalid_token: Unauthorized, invalid or missing authorization token.
  • 429 too_many_requests: Too many requests, you have exceeded the rate limit.
  • 500 api_error: Internal server error, something went wrong on the server.

Error Response Example

{
  "success": false,
  "error": {
    "code": "api_error",
    "message": "fetch failed"
  },
  "trace_id": "2cf86e86-22a4-46e1-ac2f-032c0f2a4e89"
}

Conclusion

Through this document, you have learned how to easily implement the conversational features of the official OpenAI ChatGPT using the OpenAI Chat Completion API. We hope this document helps you better integrate and use the API. If you have any questions, please feel free to contact our technical support team.