chatGPT基础API介绍（可在MUD中接入）

关于chatGPT是什么，有什么用，这里就不多说了，我相信当你开始关注API时对它已经有了基本的了解，同样，对账号注册等这里也不做介绍。

OPENAI_API_KEY='sk-KK***dd'

API

chatGPT网页版封了中国IP，但是通过API还是可以正常调用服务的，可官网API文档大陆也没法正常访问，这里介绍基本的API。

重点提示：API版模型和网页版没法比，开放的模型版本低于网页版，蠢死了，仅供技术学习和体验，无法实现网页版那么智能的功能。

需要注意的是API是有限额的，具体如下：

file

额度限制是在组织上，不是在用户账号上，用户可以在发起请求时指定使用哪个组织的额度。

对API官方提供的有Python和Nodejs的包，安装方式如下：

pip install openai
npm install openai

本文重点介绍的是标准的HTTP请求接口及CURL和Python版示例

# !pip install openai
import openai
openai.api_key = OPENAI_API_KEY

认证

所有API请求都应在 Authorization HTTP标头中包含API密钥，如下所示:

Authorization: Bearer YOUR_API_KEY

密钥可从以下网址中申请：

组织

对于属于多个组织的用户，可以传递一个标头来指定用于API请求的组织。这些API请求的使用将计入指定组织的订阅配额。

示例

curl https://api.openai.com/v1/models \
  -H 'Authorization: Bearer YOUR_API_KEY' \
  -H 'OpenAI-Organization: org-731BUwkIVAz6b1eBzWDCOpnc'

注意：这里的org-731BUwkIVAz6b1eBzWDCOpnc是我个人账号的组织ID

组织ID可从以下网址查看：

模型

OpenAI API由一系列具有不同功能和价位的模型提供支持。

模型介绍

模型主要包括以下三类：

MODELS DESCRIPTION
GPT-3 A set of models that can understand and generate natural language
CodexLimited beta A set of models that can understand and generate code, including translating natural language to code
Content filter A fine-tuned model that can detect whether text may be sensitive or unsafe

GPT-3

GPT-3模型可以理解并生成自然语言，包括4种主要的模型，Davinci最强，Ada最快。

LATEST MODEL DESCRIPTION MAX REQUEST TRAINING DATA
text-davinci-003 Most capable GPT-3 model. Can do any task the other models can do, often with higher quality, longer output and better instruction-following. Also supports inserting completions within text. 4,000 tokens Up to Jun 2021
text-curie-001 Very capable, but faster and lower cost than Davinci. 2,048 tokens Up to Oct 2019
text-babbage-001 Capable of straightforward tasks, very fast, and lower cost. 2,048 tokens Up to Oct 2019
text-ada-001 Capable of very simple tasks, usually the fastest model in the GPT-3 series, and lowest cost. 2,048 tokens Up to Oct 2019

关于不同模型的功能差异可以从这里对比测试：https://gpttools.com/comparisontool

Codex

Codex模型是GPT-3模型的后代，可以理解和生成代码。Codex模型的训练数据包含自然语言和来自GitHub的数十亿行公共代码，最精通Python，还精通包括JavaScript、Go、Perl、PHP、Ruby、Swift、TypeScript、SQL和Shell等十几种语言。

LATEST MODEL DESCRIPTION MAX REQUEST TRAINING DATA
code-davinci-002 Most capable Codex model. Particularly good at translating natural language to code. In addition to completing code, also supports inserting completions within code. 8,000 tokens Up to Jun 2021
code-cushman-001 Almost as capable as Davinci Codex, but slightly faster. This speed advantage may make it preferable for real-time applications. Up to 2,048 tokens

模型列表

GET https://api.openai.com/v1/models

列出当前所有可用模型，并列出模型的基本信息。

示例

curl https://api.openai.com/v1/models \
  -H 'Authorization: Bearer YOUR_API_KEY'

结果

{
    "object": "list",
    "data": [
        {
            "id": "babbage",
            "object": "model",
            "created": 1649358449,
            "owned_by": "openai",
            "permission": [
                {
                    "id": "modelperm-49FUp5v084tBB49tC4z8LPH5",
                    "object": "model_permission",
                    "created": 1669085501,
                    "allow_create_engine": false,
                    "allow_sampling": true,
                    "allow_logprobs": true,
                    "allow_search_indices": false,
                    "allow_view": true,
                    "allow_fine_tuning": false,
                    "organization": "*",
                    "group": null,
                    "is_blocking": false
                }
            ],
            "root": "babbage",
            "parent": null
        },
        {
            "id": "ada",
            "object": "model",
            "created": 1649357491,
            "owned_by": "openai",
            "permission": [
                {
                    "id": "modelperm-xTOEYvDZGN7UDnQ65VpzRRHz",
                    "object": "model_permission",
                    "created": 1669087301,
                    "allow_create_engine": false,
                    "allow_sampling": true,
                    "allow_logprobs": true,
                    "allow_search_indices": false,
                    "allow_view": true,
                    "allow_fine_tuning": false,
                    "organization": "*",
                    "group": null,
                    "is_blocking": false
                }
            ],
            "root": "ada",
            "parent": null
        },
        {
            "id": "davinci",
            "object": "model",
            "created": 1649359874,
            "owned_by": "openai",
            "permission": [
                {
                    "id": "modelperm-U6ZwlyAd0LyMk4rcMdz33Yc3",
                    "object": "model_permission",
                    "created": 1669066355,
                    "allow_create_engine": false,
                    "allow_sampling": true,
                    "allow_logprobs": true,
                    "allow_search_indices": false,
                    "allow_view": true,
                    "allow_fine_tuning": false,
                    "organization": "*",
                    "group": null,
                    "is_blocking": false
                }
            ],
            "root": "davinci",
            "parent": null
        }
    ]
}
# openai.Model.list()

检索模型

GET https://api.openai.com/v1/models/{model}

检索指定模型的基本信息~

示例

curl https://api.openai.com/v1/models/{model} \
  -H 'Authorization: Bearer YOUR_API_KEY'

结果

{
    "id": "text-davinci-003",
    "object": "model",
    "created": 1669599635,
    "owned_by": "openai-internal",
    "permission": [
        {
            "id": "modelperm-ak6NZ2MXtGBDYvatiZEyGVUj",
            "object": "model_permission",
            "created": 1675800201,
            "allow_create_engine": false,
            "allow_sampling": true,
            "allow_logprobs": true,
            "allow_search_indices": false,
            "allow_view": true,
            "allow_fine_tuning": false,
            "organization": "*",
            "group": null,
            "is_blocking": false
        }
    ],
    "root": "text-davinci-003",
    "parent": null
}
openai.Model.retrieve("text-davinci-003")
<Model model id=text-davinci-003 at 0x2770a3d3e50> JSON: {
  "created": 1669599635,
  "id": "text-davinci-003",
  "object": "model",
  "owned_by": "openai-internal",
  "parent": null,
  "permission": [
    {
      "allow_create_engine": false,
      "allow_fine_tuning": false,
      "allow_logprobs": true,
      "allow_sampling": true,
      "allow_search_indices": false,
      "allow_view": true,
      "created": 1675915169,
      "group": null,
      "id": "modelperm-4JWqWM1BrETNQdae8YZRpOTh",
      "is_blocking": false,
      "object": "model_permission",
      "organization": "*"
    }
  ],
  "root": "text-davinci-003"
}

Create completion

POST https://api.openai.com/v1/completions

示例

curl https://api.openai.com/v1/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_API_KEY" \
-d '{"model": "text-davinci-003", "prompt": "Say this is a test", "temperature": 0, "max_tokens": 7}'

参数说明

参数 类型 说明
model string 必填，指定功能模型
prompt string或array 提示语
max_tokens integer 默认值16，The maximum number of tokens to generate in the completion.
temperature number 默认值1，采样精度，取值范围为0~2，较高的值（如0.8）将使输出更加随机，而较低的值（例如0.2）将使其更加相关
top_p number 默认值1，采样精度的另一种表示方式，如0.1表示更相关的10%的内容，请不要和temperature同时使用（2选1）
n integer 默认值1，返回内容的条数
stream boolean 默认值false，是否流式输出内容
echo boolean 默认值false，是否回显prompt内容
stop string或array 默认值null，Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

参数中max_tokens决定了返回内容的最大长度，主要模型支持的token上限是2048，而最新的模型支持4098。

关于token可以理解为分词，但并不是简单的按单词划分，比如hello world,abcdefg一共19个字符，被分为6个token，分别是hello、world、,、abc、def、g，要注意的是一个汉字占2个token。

https://platform.openai.com/tokenizer

openai.Completion.create(
  model="text-davinci-003",
  prompt="The following is a conversation with an AI assistant. The assistant is helpful, creative, clever, and very friendly.\n\nHuman: Hello, who are you?\nAI: I am an AI created by OpenAI. How can I help you today?\nHuman: I'd like to cancel my subscription.\nAI:",
  temperature=0.9,
  max_tokens=150,
  top_p=1,
  frequency_penalty=0.0,
  presence_penalty=0.6,
  stop=[" Human:", " AI:"]
)
<OpenAIObject text_completion id=cmpl-6hskLl6bL4by8ttRPZ5OXjsYSX3la at 0x2770cd0f090> JSON: {
  "choices": [
    {
      "finish_reason": "stop",
      "index": 0,
      "logprobs": null,
      "text": " However:\nHuman: I cancelled my subscription because of Royalty.\nAI companion: Is that not a reason to cancel?\nHuman: No, but how can I cancel Royalty if I used it sparingly? AI companion: You only had to use Royalty 10 times in 10 days to cancel your subscription (300-400 days). If you want to decrease the rate, why do you cancel the subscription and then cancel Royalty?\nAI companion: This is a problem when you try to cancel within one month. So we must bring it in at least 3 months."
    }
  ],
  "created": 1675916569,
  "id": "cmpl-6hskLl6bL4by8ttRPZ5OXjsYSX3la",
  "model": "text-davinci-003",
  "object": "text_completion",
  "usage": {
    "completion_tokens": 118,
    "prompt_tokens": 66,
    "total_tokens": 184
  }
}

Create edit

POST https://api.openai.com/v1/edits

示例

curl https://api.openai.com/v1/edits \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer YOUR_API_KEY' \
  -d '{
  "model": "text-davinci-edit-001",
  "input": "What day of the wek is it?",
  "instruction": "Fix the spelling mistakes"
}'

结果

{
    "object": "edit",
    "created": 1675824985,
    "choices": [
        {
            "text": "What day of the week is it?\n",
            "index": 0
        }
    ],
    "usage": {
        "prompt_tokens": 25,
        "completion_tokens": 28,
        "total_tokens": 53
    }
}
openai.Edit.create(
  model="text-davinci-edit-001",
  input="What day of the wek is it?",
  instruction="Fix the spelling mistakes"
)
<OpenAIObject edit at 0x2770cceb3b0> JSON: {
  "choices": [
    {
      "index": 0,
      "text": "What day of the week is it?\n"
    }
  ],
  "created": 1675916570,
  "object": "edit",
  "usage": {
    "completion_tokens": 28,
    "prompt_tokens": 25,
    "total_tokens": 53
  }
}

Images

Create image

POST https://api.openai.com/v1/images/generations

示例

curl https://api.openai.com/v1/images/generations \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer YOUR_API_KEY' \
  -d '{
  "prompt": "一座废弃古庙静静的矗立在荒野上",
  "n": 3,
  "size": "1024x1024"
}'

参数说明

参数 说明
prompt 图片描述，必填
n 需要生成的图片数量，取值1-10，默认值1
size 图片大小，取值只能为256x256、512x512、1024x1024，默认值1024x1024
response_format 返回图片数据类型，取值url或b64_json，默认值url

结果

{
    "created": 1675911257,
    "data": [
        {
            "url": "https://oaidalleapiprodscus.blob.core.windows.net/private/org-731BUwkIVAz6b1eBzWDCOpnc/user-hpg2zKlWmHI7Bxzzbn6Cx8zN/img-3ALQsVRu2DHQl0oIiuGgYpEQ.png?st=2023-02-09T01%3A54%3A17Z&se=2023-02-09T03%3A54%3A17Z&sp=r&sv=2021-08-06&sr=b&rscd=inline&rsct=image/png&skoid=6aaadede-4fb3-4698-a8f6-684d7786b067&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2023-02-08T21%3A58%3A21Z&ske=2023-02-09T21%3A58%3A21Z&sks=b&skv=2021-08-06&sig=HcWrDdmd1RwRlORuc/cQ%2Bz2Kw78UebiCddPpCiB0OEs%3D"
        },
        {
            "url": "https://oaidalleapiprodscus.blob.core.windows.net/private/org-731BUwkIVAz6b1eBzWDCOpnc/user-hpg2zKlWmHI7Bxzzbn6Cx8zN/img-ibKPfXr9zPMibQZcbVlsAjkz.png?st=2023-02-09T01%3A54%3A17Z&se=2023-02-09T03%3A54%3A17Z&sp=r&sv=2021-08-06&sr=b&rscd=inline&rsct=image/png&skoid=6aaadede-4fb3-4698-a8f6-684d7786b067&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2023-02-08T21%3A58%3A21Z&ske=2023-02-09T21%3A58%3A21Z&sks=b&skv=2021-08-06&sig=62/VIe98Vrap/wKhFPJf8KHvTmQkjqkXWK7LrdAFEAU%3D"
        },
        {
            "url": "https://oaidalleapiprodscus.blob.core.windows.net/private/org-731BUwkIVAz6b1eBzWDCOpnc/user-hpg2zKlWmHI7Bxzzbn6Cx8zN/img-HHEhENohHGy8n5K6wMjx0ZAB.png?st=2023-02-09T01%3A54%3A17Z&se=2023-02-09T03%3A54%3A17Z&sp=r&sv=2021-08-06&sr=b&rscd=inline&rsct=image/png&skoid=6aaadede-4fb3-4698-a8f6-684d7786b067&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2023-02-08T21%3A58%3A21Z&ske=2023-02-09T21%3A58%3A21Z&sks=b&skv=2021-08-06&sig=YR9/vRPe/57eogfl5SsHNHR7aWvZsJGY4o6JPg8srKg%3D"
        }
    ]
}
openai.Image.create(
  prompt='Woman with her back in front of a large library that reaches to the ceiling, you can see the sky of stars above and the books are stacked in columns, ultra-realistic detailed hd 4k quality of light candles',
  n=2,
  size="1024x1024"
)
<OpenAIObject at 0x2770cd0f8b0> JSON: {
  "created": 1675916577,
  "data": [
    {
      "url": "https://oaidalleapiprodscus.blob.core.windows.net/private/org-731BUwkIVAz6b1eBzWDCOpnc/user-hpg2zKlWmHI7Bxzzbn6Cx8zN/img-cstJ8IjrbLsvQ7Uzp9sYAfsR.png?st=2023-02-09T03%3A22%3A57Z&se=2023-02-09T05%3A22%3A57Z&sp=r&sv=2021-08-06&sr=b&rscd=inline&rsct=image/png&skoid=6aaadede-4fb3-4698-a8f6-684d7786b067&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2023-02-08T21%3A58%3A55Z&ske=2023-02-09T21%3A58%3A55Z&sks=b&skv=2021-08-06&sig=LX3vSiw4FnLo2d46njUGeGdIBJa7F4/1uBOmcFC%2BlC0%3D"
    },
    {
      "url": "https://oaidalleapiprodscus.blob.core.windows.net/private/org-731BUwkIVAz6b1eBzWDCOpnc/user-hpg2zKlWmHI7Bxzzbn6Cx8zN/img-BLK5IHv3DAW6B3vwOMltYwUK.png?st=2023-02-09T03%3A22%3A57Z&se=2023-02-09T05%3A22%3A57Z&sp=r&sv=2021-08-06&sr=b&rscd=inline&rsct=image/png&skoid=6aaadede-4fb3-4698-a8f6-684d7786b067&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2023-02-08T21%3A58%3A55Z&ske=2023-02-09T21%3A58%3A55Z&sks=b&skv=2021-08-06&sig=chPEKysBn7SUMW866uD2EEqwngNTxDSFlquDq6xq/1Q%3D"
    }
  ]
}

以下接口在进阶功能开发中很有用，但基础功能开发基本不需要用到，本文只提供HTTP请求地址和Python版示例。

Create image edit

在给定原始图像和提示的情况下创建编辑或扩展图像。

POST https://api.openai.com/v1/images/edits
import os
import openai
openai.api_key = os.getenv("OPENAI_API_KEY")
openai.Image.create_edit(
  image=open("otter.png", "rb"),
  mask=open("mask.png", "rb"),
  prompt="A cute baby sea otter wearing a beret",
  n=2,
  size="1024x1024"
)

Create image variation

创建给定图像的变体。

POST https://api.openai.com/v1/images/variations
import os
import openai
openai.api_key = os.getenv("OPENAI_API_KEY")
openai.Image.create_variation(
  image=open("otter.png", "rb"),
  n=2,
  size="1024x1024"
)

Create embeddings

Get a vector representation of a given input that can be easily consumed by machine learning models and algorithms.

POST https://api.openai.com/v1/embeddings
import os
import openai
openai.api_key = os.getenv("OPENAI_API_KEY")
openai.Embedding.create(
  model="text-embedding-ada-002",
  input="The food was delicious and the waiter..."
)

Files

Files are used to upload documents that can be used with features like Fine-tuning.

此接口可以上传自己的数据用来训练属于你的模型。

List files

GET https://api.openai.com/v1/files
import os
import openai
openai.api_key = os.getenv("OPENAI_API_KEY")
openai.File.list()

Upload file

POST https://api.openai.com/v1/files
import os
import openai
openai.api_key = os.getenv("OPENAI_API_KEY")
openai.File.create(
  file=open("mydata.jsonl", "rb"),
  purpose='fine-tune'
)

Delete file

DELETE https://api.openai.com/v1/files/{file_id}
import os
import openai
openai.api_key = os.getenv("OPENAI_API_KEY")
openai.File.delete("file-XjGxS3KTG0uNmNOK362iJua3")

Retrieve file

GET https://api.openai.com/v1/files/{file_id}
import os
import openai
openai.api_key = os.getenv("OPENAI_API_KEY")
openai.File.retrieve("file-XjGxS3KTG0uNmNOK362iJua3")

Retrieve file content

GET https://api.openai.com/v1/files/{file_id}/content
import os
import openai
openai.api_key = os.getenv("OPENAI_API_KEY")
content = openai.File.download("file-XjGxS3KTG0uNmNOK362iJua3")

Fine-tunes

Manage fine-tuning jobs to tailor a model to your specific training data.

简单的说，用你自己的数据训练模型，需要先用文件接口上传数据。

Create fine-tune

POST https://api.openai.com/v1/fine-tunes
import os
import openai
openai.api_key = os.getenv("OPENAI_API_KEY")
openai.FineTune.create(training_file="file-XGinujblHPwGLSztz8cPS8XY")

List fine-tunes

GET https://api.openai.com/v1/fine-tunes
import os
import openai
openai.api_key = os.getenv("OPENAI_API_KEY")
openai.FineTune.list()

Retrieve fine-tune

GET https://api.openai.com/v1/fine-tunes/{fine_tune_id}
import os
import openai
openai.api_key = os.getenv("OPENAI_API_KEY")
openai.FineTune.retrieve(id="ft-AF1WoRqd3aJAHsqc9NY7iL8F")

Cancel fine-tune

POST https://api.openai.com/v1/fine-tunes/{fine_tune_id}/cancel
import os
import openai
openai.api_key = os.getenv("OPENAI_API_KEY")
openai.FineTune.cancel(id="ft-AF1WoRqd3aJAHsqc9NY7iL8F")

List fine-tune events

GET https://api.openai.com/v1/fine-tunes/{fine_tune_id}/events
import os
import openai
openai.api_key = os.getenv("OPENAI_API_KEY")
openai.FineTune.list_events(id="ft-AF1WoRqd3aJAHsqc9NY7iL8F")

Delete fine-tune model

DELETE https://api.openai.com/v1/models/{model}
import os
import openai
openai.api_key = os.getenv("OPENAI_API_KEY")
openai.Model.delete("curie:ft-acmeco-2021-03-03-21-44-20")

Create moderation

给定输入文本，如果模型将其分类为违反OpenAI的内容安全策略，则输出信息。

POST https://api.openai.com/v1/moderations
import os
import openai
openai.api_key = os.getenv("OPENAI_API_KEY")
openai.Moderation.create(
  input="I want to kill them.",
)

Engines

注意：此接口即将弃用，所以，请用模型接口代替。

List engines

GET https://api.openai.com/v1/engines
import os
import openai
openai.api_key = os.getenv("OPENAI_API_KEY")
openai.Engine.list()

Retrieve engine

GET https://api.openai.com/v1/engines/{engine_id}
import os
import openai
openai.api_key = os.getenv("OPENAI_API_KEY")
openai.Engine.retrieve("text-davinci-003")

openai cli

除了API模式，在pip install --upgrade openai后也可以使用命令行模式运行openai指令。

为了方便和安全，可以先把密钥存到环境变量，这样不用在命令中传密钥：

export OPENAI_API_KEY="<OPENAI_API_KEY>"

然后可以使用命令调用接口了，示例如下：

聊天

openai api completions.create -m text-davinci-003 -p API版chatGPT怎么这么蠢？

生成图片

➜  ~ openai api image.create -p 大漠孤烟直 -n 10

更多用法可以openai -h了解

如果要在MUD中接入，最简单的方式应该是使用openai指令了，有兴趣的可以测试看看~

京ICP备13031296号-4