文件输入 | OpenAI API

文件输入的支持情况取决于 API 端点。Responses API 接受下列文件类型作为 input_file 项。Chat Completions 仅接受 PDF 文件作为 file 内容部分。

输入方式	Responses API	Chat Completions
Base64 编码的文件数据（`file_data`）	下列支持的文件类型	仅限 PDF
已上传文件的 ID（`file_id`）	下列支持的文件类型	仅限 PDF
外部文件 URL（`file_url`）	下列支持的文件类型	不支持

对于非 PDF 文件输入，请使用 Responses API。要在 Chat Completions 中使用文件中的文本，请在您的应用中读取文件，并将其内容作为 text 内容部分发送。

工作原理

在 Responses API 中，input_file 的处理方式取决于文件类型：

PDF 文件：对于具有视觉能力的模型（例如 gpt-4o 及更新的模型），API 会提取文本和页面图像，并将两者一并发送给模型。
非 PDF 文档和文本文件 （例如 .docx、.pptx、.txt 和代码文件）：API 仅提取文本。
电子表格文件 （例如 .xlsx、.csv、.tsv）：API 会运行专门针对电子表格的增强流程（详见下文）。

如果以下相关工具更适合您的任务，请使用它们：

使用文件搜索来检索大型文件，而不是将这些文件直接作为 input_file 传入。
对于需要详细分析且涉及大量电子表格处理的任务，例如聚合、连接、绘图或自定义计算，请使用托管式 Shell。

非 PDF 文件的图像和图表限制

对于非 PDF 文件，Responses API 不会将其中嵌入的图像或图表提取到模型上下文中。

为保留图表和示意图的细节，请先将文件转换为 PDF，然后将 PDF 作为 input_file 发送。

电子表格增强的工作原理

对于电子表格类文件（例如 .xlsx、.xls、.csv、.tsv 和 .iif），Responses API 会采用专门针对电子表格的增强流程。

API 不会将整个工作表传给模型，而是最多解析每个工作表的前 1,000 行，并添加模型生成的摘要和表头元数据，让模型能够基于更精简的结构化数据视图开展工作。

PDF 细节级别

对于 Responses API 中的 PDF 输入，可将 input_file 项中的可选 detail 字段设为 auto、low 或 high，以控制 API 处理页面图像的方式。如果省略，detail 默认为 auto。对于 GPT-5.6 及更新的模型，auto 使用 high；对于更早的模型，则使用 low。使用 low 可减少输入 Token 用量；使用 high 则可获取更多视觉细节，例如密集图表、小字号文字或示意图中的细节。

detail 设置仅影响 PDF 页面图像的处理。从 PDF 中提取的文本仍会包含在内。Chat Completions 的文件输入不支持 detail。

以下是一个明确指定高细节级别的最简 Responses API 请求体：

1234567891011121314151617181920{
  "model": "gpt-4.1",
  "input": [
    {
      "role": "user",
      "content": [
        {
          "type": "input_file",
          "filename": "document.pdf",
          "file_data": "data:application/pdf;base64,...",
          "detail": "high"
        },
        {
          "type": "input_text",
          "text": "Summarize this document."
        }
      ]
    }
  ]
}

支持的文件类型

下表列出了 Responses API 接受作为 input_file 项的常见文件类型。扩展名和 MIME 类型的完整列表见本页后文。Chat Completions 仅支持 .pdf（application/pdf），无论使用 file_data 还是 file_id。

类别	常见扩展名
PDF 文件	`.pdf`
文本和代码	`.txt`、`.md`、`.json`、`.html`、`.xml`、代码文件
富文本文档	`.doc`、`.docx`、`.rtf`、`.odt`
演示文稿	`.ppt`、`.pptx`
电子表格	`.csv`、`.xls`、`.xlsx`

文件 URL

您可以通过外部 URL 链接提供文件输入。

使用外部文件 URL

curl

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23import OpenAI from "openai";
const client = new OpenAI();

const response = await client.responses.create({
  model: "gpt-6-astra",
  input: [
    {
      role: "user",
      content: [
        {
          type: "input_text",
          text: "Analyze the letter and provide a summary of the key points.",
        },
        {
          type: "input_file",
          file_url: "https://www.berkshirehathaway.com/letters/2024ltr.pdf",
        },
      ],
    },
  ],
});

console.log(response.output_text);

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="gpt-6-astra",
    input=[
        {
            "role": "user",
            "content": [
                {
                    "type": "input_text",
                    "text": "Analyze the letter and provide a summary of the key points.",
                },
                {
                    "type": "input_file",
                    "file_url": "https://www.berkshirehathaway.com/letters/2024ltr.pdf",
                },
            ],
        },
    ],
)

print(response.output_text)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41package main

import (
	"context"
	"fmt"

	"github.com/openai/openai-go/v3"
	"github.com/openai/openai-go/v3/responses"
)

func main() {
	client := openai.NewClient()

	response, err := client.Responses.New(context.Background(), responses.ResponseNewParams{
		Model: "gpt-6-astra",
		Input: responses.ResponseNewParamsInputUnion{
			OfInputItemList: responses.ResponseInputParam{
				responses.ResponseInputItemParamOfMessage(
					responses.ResponseInputMessageContentListParam{
						responses.ResponseInputContentParamOfInputText(
							"Analyze the letter and provide a summary of the key points.",
						),
						{
							OfInputFile: &responses.ResponseInputFileParam{
								FileURL: openai.String(
									"https://www.berkshirehathaway.com/letters/2024ltr.pdf",
								),
							},
						},
					},
					responses.EasyInputMessageRoleUser,
				),
			},
		},
	})
	if err != nil {
		panic(err)
	}

	fmt.Println(response.OutputText())
}

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30import com.openai.client.OpenAIClient;
import com.openai.client.okhttp.OpenAIOkHttpClient;
import com.openai.models.responses.ResponseCreateParams;
import com.openai.models.responses.ResponseInputFile;
import com.openai.models.responses.ResponseInputItem;
import java.util.List;

ResponseCreateParams params =
    ResponseCreateParams.builder()
        .model("gpt-6-astra")
        .inputOfResponse(
            List.of(
                ResponseInputItem.ofMessage(
                    ResponseInputItem.Message.builder()
                        .role(ResponseInputItem.Message.Role.USER)
                        .addInputTextContent(
                            "Analyze the letter and provide a summary of the key points.")
                        .addContent(
                            ResponseInputFile.builder()
                                .fileUrl(
                                    "https://www.berkshirehathaway.com/letters/2024ltr.pdf")
                                .build())
                        .build())))
        .build();

client.responses().create(params).output().stream()
    .flatMap(item -> item.message().stream())
    .flatMap(message -> message.content().stream())
    .flatMap(content -> content.outputText().stream())
    .forEach(text -> System.out.println(text.text()));

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25using OpenAI.Responses;
#pragma warning disable OPENAI001

string key = Environment.GetEnvironmentVariable("OPENAI_API_KEY")!;
ResponsesClient client = new(key);

Uri fileUrl = new(
    "https://www.berkshirehathaway.com/letters/2024ltr.pdf"
);

ResponseResult response = await client.CreateResponseAsync(
    "gpt-6-astra",
    [
        ResponseItem.CreateUserMessageItem(
            [
                ResponseContentPart.CreateInputTextPart(
                    "Analyze the letter and provide a summary of the key points."
                ),
                ResponseContentPart.CreateInputFilePart(fileUrl),
            ]
        ),
    ]
);

Console.WriteLine(response.GetOutputText());

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24require "openai"

openai = OpenAI::Client.new

response = openai.responses.create(
  model: "gpt-6-astra",
  input: [
    {
      role: "user",
      content: [
        {
          type: "input_text",
          text: "Analyze the letter and provide a summary of the key points."
        },
        {
          type: "input_file",
          file_url: "https://www.berkshirehathaway.com/letters/2024ltr.pdf"
        }
      ]
    }
  ]
)

puts(response.output_text)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21curl "https://api.openai.com/v1/responses" \
    -H "Content-Type: application/json" \
    -H "Authorization: Bearer $OPENAI_API_KEY" \
    -d '{
        "model": "gpt-6-astra",
        "input": [
            {
                "role": "user",
                "content": [
                    {
                        "type": "input_text",
                        "text": "Analyze the letter and provide a summary of the key points."
                    },
                    {
                        "type": "input_file",
                        "file_url": "https://www.berkshirehathaway.com/letters/2024ltr.pdf"
                    }
                ]
            }
        ]
    }'

上传文件

以下示例使用 Files API 上传文件，然后在发送给模型的请求中引用该文件的 ID。

上传文件

curl

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29import fs from "fs";
import OpenAI from "openai";
const client = new OpenAI();

const file = await client.files.create({
  file: fs.createReadStream("fixtures/draconomicon.pdf"),
  purpose: "user_data",
});

const response = await client.responses.create({
  model: "gpt-6-astra",
  input: [
    {
      role: "user",
      content: [
        {
          type: "input_file",
          file_id: file.id,
        },
        {
          type: "input_text",
          text: "What is the first dragon in the book?",
        },
      ],
    },
  ],
});

console.log(response.output_text);

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26from openai import OpenAI

client = OpenAI()

file = client.files.create(file=open("draconomicon.pdf", "rb"), purpose="user_data")

response = client.responses.create(
    model="gpt-6-astra",
    input=[
        {
            "role": "user",
            "content": [
                {
                    "type": "input_file",
                    "file_id": file.id,
                },
                {
                    "type": "input_text",
                    "text": "What is the first dragon in the book?",
                },
            ],
        }
    ],
)

print(response.output_text)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54package main

import (
	"context"
	"fmt"
	"os"

	"github.com/openai/openai-go/v3"
	"github.com/openai/openai-go/v3/responses"
)

func main() {
	client := openai.NewClient()

	file, err := os.Open("draconomicon.pdf")
	if err != nil {
		panic(err)
	}
	defer file.Close()

	uploadedFile, err := client.Files.New(context.Background(), openai.FileNewParams{
		File:    file,
		Purpose: openai.FilePurposeUserData,
	})
	if err != nil {
		panic(err)
	}

	response, err := client.Responses.New(context.Background(), responses.ResponseNewParams{
		Model: "gpt-6-astra",
		Input: responses.ResponseNewParamsInputUnion{
			OfInputItemList: responses.ResponseInputParam{
				responses.ResponseInputItemParamOfMessage(
					responses.ResponseInputMessageContentListParam{
						{
							OfInputFile: &responses.ResponseInputFileParam{
								FileID: openai.String(uploadedFile.ID),
							},
						},
						responses.ResponseInputContentParamOfInputText(
							"What is the first dragon in the book?",
						),
					},
					responses.EasyInputMessageRoleUser,
				),
			},
		},
	})
	if err != nil {
		panic(err)
	}

	fmt.Println(response.OutputText())
}

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40import com.openai.client.OpenAIClient;
import com.openai.client.okhttp.OpenAIOkHttpClient;
import com.openai.models.files.FileCreateParams;
import com.openai.models.files.FilePurpose;
import com.openai.models.responses.ResponseCreateParams;
import com.openai.models.responses.ResponseInputFile;
import com.openai.models.responses.ResponseInputItem;
import java.nio.file.Path;
import java.util.List;

var file =
    client
        .files()
        .create(
            FileCreateParams.builder()
                .file(Path.of(System.getenv("OPENAI_EXAMPLE_FILE_PATH")))
                .purpose(FilePurpose.USER_DATA)
                .build());

var response =
    client
        .responses()
        .create(
            ResponseCreateParams.builder()
                .model("gpt-6-astra")
                .inputOfResponse(
                    List.of(
                        ResponseInputItem.ofMessage(
                            ResponseInputItem.Message.builder()
                                .role(ResponseInputItem.Message.Role.USER)
                                .addContent(
                                    ResponseInputFile.builder().fileId(file.id()).build())
                                .addInputTextContent("What is the first dragon in the book?")
                                .build())))
                .build());
response.output().stream()
    .flatMap(item -> item.message().stream())
    .flatMap(message -> message.content().stream())
    .flatMap(content -> content.outputText().stream())
    .forEach(text -> System.out.println(text.text()));

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29using OpenAI.Files;
using OpenAI.Responses;
#pragma warning disable OPENAI001

string key = Environment.GetEnvironmentVariable("OPENAI_API_KEY")!;
ResponsesClient client = new(key);

OpenAIFileClient files = new(key);

OpenAIFile file = await files.UploadFileAsync(
    "draconomicon.pdf",
    FileUploadPurpose.UserData
);

ResponseResult response = await client.CreateResponseAsync(
    "gpt-6-astra",
    [
        ResponseItem.CreateUserMessageItem(
            [
                ResponseContentPart.CreateInputFilePart(file.Id),
                ResponseContentPart.CreateInputTextPart(
                    "What is the first dragon in the book?"
                ),
            ]
        ),
    ]
);

Console.WriteLine(response.GetOutputText());

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30require "openai"
require "pathname"

openai = OpenAI::Client.new

file = openai.files.create(
  file: Pathname("draconomicon.pdf"),
  purpose: "user_data"
)

response = openai.responses.create(
  model: "gpt-6-astra",
  input: [
    {
      role: "user",
      content: [
        {
          type: "input_file",
          file_id: file.id
        },
        {
          type: "input_text",
          text: "What is the first dragon in the book?"
        }
      ]
    }
  ]
)

puts(response.output_text)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26curl https://api.openai.com/v1/files \
    -H "Authorization: Bearer $OPENAI_API_KEY" \
    -F purpose="user_data" \
    -F file="@draconomicon.pdf"

curl "https://api.openai.com/v1/responses" \
    -H "Content-Type: application/json" \
    -H "Authorization: Bearer $OPENAI_API_KEY" \
    -d '{
        "model": "gpt-6-astra",
        "input": [
            {
                "role": "user",
                "content": [
                    {
                        "type": "input_file",
                        "file_id": "file-6F2ksmvXxt4VdoqmHRw6kL"
                    },
                    {
                        "type": "input_text",
                        "text": "What is the first dragon in the book?"
                    }
                ]
            }
        ]
    }'

上传文件

curl

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31import fs from "fs";
import OpenAI from "openai";
const client = new OpenAI();

const file = await client.files.create({
  file: fs.createReadStream("fixtures/draconomicon.pdf"),
  purpose: "user_data",
});

const completion = await client.chat.completions.create({
  model: "gpt-6-astra",
  messages: [
    {
      role: "user",
      content: [
        {
          type: "file",
          file: {
            file_id: file.id,
          },
        },
        {
          type: "text",
          text: "What is the first dragon in the book?",
        },
      ],
    },
  ],
});

console.log(completion.choices[0].message.content);

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28from openai import OpenAI

client = OpenAI()

file = client.files.create(file=open("draconomicon.pdf", "rb"), purpose="user_data")

completion = client.chat.completions.create(
    model="gpt-6-astra",
    messages=[
        {
            "role": "user",
            "content": [
                {
                    "type": "file",
                    "file": {
                        "file_id": file.id,
                    },
                },
                {
                    "type": "text",
                    "text": "What is the first dragon in the book?",
                },
            ],
        }
    ],
)

print(completion.choices[0].message.content)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44package main

import (
	"context"
	"fmt"
	"os"

	"github.com/openai/openai-go/v3"
)

func main() {
	client := openai.NewClient()

	file, err := os.Open("draconomicon.pdf")
	if err != nil {
		panic(err)
	}
	defer file.Close()

	uploadedFile, err := client.Files.New(context.Background(), openai.FileNewParams{
		File:    file,
		Purpose: openai.FilePurposeUserData,
	})
	if err != nil {
		panic(err)
	}

	completion, err := client.Chat.Completions.New(context.Background(), openai.ChatCompletionNewParams{
		Model: "gpt-6-astra",
		Messages: []openai.ChatCompletionMessageParamUnion{
			openai.UserMessage([]openai.ChatCompletionContentPartUnionParam{
				openai.FileContentPart(openai.ChatCompletionContentPartFileFileParam{
					FileID: openai.String(uploadedFile.ID),
				}),
				openai.TextContentPart("What is the first dragon in the book?"),
			}),
		},
	})
	if err != nil {
		panic(err)
	}

	fmt.Println(completion.Choices[0].Message.Content)
}

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43import com.openai.client.OpenAIClient;
import com.openai.client.okhttp.OpenAIOkHttpClient;
import com.openai.models.chat.completions.ChatCompletionContentPart;
import com.openai.models.chat.completions.ChatCompletionContentPartText;
import com.openai.models.chat.completions.ChatCompletionCreateParams;
import com.openai.models.chat.completions.ChatCompletionUserMessageParam;
import com.openai.models.files.FileCreateParams;
import com.openai.models.files.FilePurpose;
import java.nio.file.Path;
import java.util.List;

var file =
    client
        .files()
        .create(
            FileCreateParams.builder()
                .file(Path.of(System.getenv("OPENAI_EXAMPLE_FILE_PATH")))
                .purpose(FilePurpose.USER_DATA)
                .build());

var filePart =
    ChatCompletionContentPart.ofFile(
        ChatCompletionContentPart.File.builder()
            .file(ChatCompletionContentPart.File.FileObject.builder().fileId(file.id()).build())
            .build());
var textPart =
    ChatCompletionContentPart.ofText(
        ChatCompletionContentPartText.builder().text("Summarize this PDF.").build());
var completion =
    client
        .chat()
        .completions()
        .create(
            ChatCompletionCreateParams.builder()
                .model("gpt-6-astra")
                .addMessage(
                    ChatCompletionUserMessageParam.builder()
                        .contentOfArrayOfContentParts(List.of(filePart, textPart))
                        .build())
                .build());
completion.choices().stream()
    .flatMap(choice -> choice.message().content().stream())
    .forEach(System.out::println);

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22using OpenAI.Chat;
using OpenAI.Files;
#pragma warning disable OPENAI001

string key = Environment.GetEnvironmentVariable("OPENAI_API_KEY")!;
string model = "gpt-6-astra";
ChatClient client = new(model, key);

OpenAIFileClient files = new(key);

OpenAIFile file = await files.UploadFileAsync("draconomicon.pdf", FileUploadPurpose.UserData);

ChatCompletion completion = await client.CompleteChatAsync(
    new UserChatMessage(
        [
            ChatMessageContentPart.CreateFilePart(file.Id),
            ChatMessageContentPart.CreateTextPart("What is the first dragon in the book?"),
        ]
    )
);

Console.WriteLine(completion.Content[0].Text);

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28require "openai"
require "pathname"

client = OpenAI::Client.new
file = client.files.create(
  file: Pathname("draconomicon.pdf"),
  purpose: :user_data
)
completion = client.chat.completions.create(
  model: "gpt-6-astra",
  messages: [
    {
      role: :user,
      content: [
        {
          type: :file,
          file: { file_id: file.id }
        },
        {
          type: :text,
          text: "Summarize this PDF."
        }
      ]
    }
  ]
)

puts(completion.choices.fetch(0).message.content)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28curl https://api.openai.com/v1/files \
    -H "Authorization: Bearer $OPENAI_API_KEY" \
    -F purpose="user_data" \
    -F file="@draconomicon.pdf"

curl "https://api.openai.com/v1/chat/completions" \
    -H "Content-Type: application/json" \
    -H "Authorization: Bearer $OPENAI_API_KEY" \
    -d '{
        "model": "gpt-6-astra",
        "messages": [
            {
                "role": "user",
                "content": [
                    {
                        "type": "file",
                        "file": {
                            "file_id": "file-6F2ksmvXxt4VdoqmHRw6kL"
                        }
                    },
                    {
                        "type": "text",
                        "text": "What is the first dragon in the book?"
                    }
                ]
            }
        ]
    }'

Base64 编码文件

您也可以将文件输入作为 Base64 编码的文件数据发送。

发送 Base64 编码文件

curl

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28import fs from "fs";
import OpenAI from "openai";
const client = new OpenAI();

const data = fs.readFileSync("fixtures/draconomicon.pdf");
const base64String = data.toString("base64");

const response = await client.responses.create({
  model: "gpt-6-astra",
  input: [
    {
      role: "user",
      content: [
        {
          type: "input_file",
          filename: "draconomicon.pdf",
          file_data: `data:application/pdf;base64,${base64String}`,
        },
        {
          type: "input_text",
          text: "What is the first dragon in the book?",
        },
      ],
    },
  ],
});

console.log(response.output_text);

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31import base64
from openai import OpenAI

client = OpenAI()

with open("draconomicon.pdf", "rb") as f:
    data = f.read()

base64_string = base64.b64encode(data).decode("utf-8")

response = client.responses.create(
    model="gpt-6-astra",
    input=[
        {
            "role": "user",
            "content": [
                {
                    "type": "input_file",
                    "filename": "draconomicon.pdf",
                    "file_data": f"data:application/pdf;base64,{base64_string}",
                },
                {
                    "type": "input_text",
                    "text": "What is the first dragon in the book?",
                },
            ],
        },
    ],
)

print(response.output_text)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48package main

import (
	"context"
	"encoding/base64"
	"fmt"
	"os"

	"github.com/openai/openai-go/v3"
	"github.com/openai/openai-go/v3/responses"
)

func main() {
	client := openai.NewClient()

	data, err := os.ReadFile("draconomicon.pdf")
	if err != nil {
		panic(err)
	}
	fileData := "data:application/pdf;base64," + base64.StdEncoding.EncodeToString(data)

	response, err := client.Responses.New(context.Background(), responses.ResponseNewParams{
		Model: "gpt-6-astra",
		Input: responses.ResponseNewParamsInputUnion{
			OfInputItemList: responses.ResponseInputParam{
				responses.ResponseInputItemParamOfMessage(
					responses.ResponseInputMessageContentListParam{
						{
							OfInputFile: &responses.ResponseInputFileParam{
								Filename: openai.String("draconomicon.pdf"),
								FileData: openai.String(fileData),
							},
						},
						responses.ResponseInputContentParamOfInputText(
							"What is the first dragon in the book?",
						),
					},
					responses.EasyInputMessageRoleUser,
				),
			},
		},
	})
	if err != nil {
		panic(err)
	}

	fmt.Println(response.OutputText())
}

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36import com.openai.client.OpenAIClient;
import com.openai.client.okhttp.OpenAIOkHttpClient;
import com.openai.models.responses.ResponseCreateParams;
import com.openai.models.responses.ResponseInputFile;
import com.openai.models.responses.ResponseInputItem;
import java.io.IOException;
import java.nio.file.Files;
import java.nio.file.Path;
import java.util.Base64;
import java.util.List;

String pdfData =
    Base64.getEncoder()
        .encodeToString(Files.readAllBytes(Path.of(System.getenv("OPENAI_EXAMPLE_FILE_PATH"))));
ResponseCreateParams params =
    ResponseCreateParams.builder()
        .model("gpt-6-astra")
        .inputOfResponse(
            List.of(
                ResponseInputItem.ofMessage(
                    ResponseInputItem.Message.builder()
                        .role(ResponseInputItem.Message.Role.USER)
                        .addContent(
                            ResponseInputFile.builder()
                                .filename("document.pdf")
                                .fileData("data:application/pdf;base64," + pdfData)
                                .build())
                        .addInputTextContent("Summarize this document.")
                        .build())))
        .build();

client.responses().create(params).output().stream()
    .flatMap(item -> item.message().stream())
    .flatMap(message -> message.content().stream())
    .flatMap(content -> content.outputText().stream())
    .forEach(text -> System.out.println(text.text()));

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26using OpenAI.Responses;
#pragma warning disable OPENAI001

string key = Environment.GetEnvironmentVariable("OPENAI_API_KEY")!;
ResponsesClient client = new(key);

BinaryData fileBytes = BinaryData.FromBytes(await File.ReadAllBytesAsync("draconomicon.pdf"));
ResponseResult response = await client.CreateResponseAsync(
    "gpt-6-astra",
    [
        ResponseItem.CreateUserMessageItem(
            [
                ResponseContentPart.CreateInputFilePart(
                    fileBytes,
                    "application/pdf",
                    "draconomicon.pdf"
                ),
                ResponseContentPart.CreateInputTextPart(
                    "What is the first dragon in the book?"
                ),
            ]
        ),
    ]
);

Console.WriteLine(response.GetOutputText());

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26require "base64"
require "openai"

client = OpenAI::Client.new
pdf_data = Base64.strict_encode64(File.binread("draconomicon.pdf"))
response = client.responses.create(
  model: "gpt-6-astra",
  input: [
    {
      role: :user,
      content: [
        {
          type: :input_file,
          filename: "document.pdf",
          file_data: "data:application/pdf;base64,#{pdf_data}"
        },
        {
          type: :input_text,
          text: "Summarize this document."
        }
      ]
    }
  ]
)

puts(response.output_text)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22curl "https://api.openai.com/v1/responses" \
    -H "Content-Type: application/json" \
    -H "Authorization: Bearer $OPENAI_API_KEY" \
    -d '{
        "model": "gpt-6-astra",
        "input": [
            {
                "role": "user",
                "content": [
                    {
                        "type": "input_file",
                        "filename": "draconomicon.pdf",
                        "file_data": "...base64 encoded PDF bytes here..."
                    },
                    {
                        "type": "input_text",
                        "text": "What is the first dragon in the book?"
                    }
                ]
            }
        ]
    }'

发送 Base64 编码文件

curl

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30import fs from "fs";
import OpenAI from "openai";
const client = new OpenAI();

const data = fs.readFileSync("fixtures/draconomicon.pdf");
const base64String = data.toString("base64");

const completion = await client.chat.completions.create({
  model: "gpt-6-astra",
  messages: [
    {
      role: "user",
      content: [
        {
          type: "file",
          file: {
            filename: "draconomicon.pdf",
            file_data: `data:application/pdf;base64,${base64String}`,
          },
        },
        {
          type: "text",
          text: "What is the first dragon in the book?",
        },
      ],
    },
  ],
});

console.log(completion.choices[0].message.content);

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33import base64
from openai import OpenAI

client = OpenAI()

with open("draconomicon.pdf", "rb") as f:
    data = f.read()

base64_string = base64.b64encode(data).decode("utf-8")

completion = client.chat.completions.create(
    model="gpt-6-astra",
    messages=[
        {
            "role": "user",
            "content": [
                {
                    "type": "file",
                    "file": {
                        "filename": "draconomicon.pdf",
                        "file_data": f"data:application/pdf;base64,{base64_string}",
                    },
                },
                {
                    "type": "text",
                    "text": "What is the first dragon in the book?",
                },
            ],
        },
    ],
)

print(completion.choices[0].message.content)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38package main

import (
	"context"
	"encoding/base64"
	"fmt"
	"os"

	"github.com/openai/openai-go/v3"
)

func main() {
	client := openai.NewClient()

	data, err := os.ReadFile("draconomicon.pdf")
	if err != nil {
		panic(err)
	}
	fileData := "data:application/pdf;base64," + base64.StdEncoding.EncodeToString(data)

	completion, err := client.Chat.Completions.New(context.Background(), openai.ChatCompletionNewParams{
		Model: "gpt-6-astra",
		Messages: []openai.ChatCompletionMessageParamUnion{
			openai.UserMessage([]openai.ChatCompletionContentPartUnionParam{
				openai.FileContentPart(openai.ChatCompletionContentPartFileFileParam{
					Filename: openai.String("draconomicon.pdf"),
					FileData: openai.String(fileData),
				}),
				openai.TextContentPart("What is the first dragon in the book?"),
			}),
		},
	})
	if err != nil {
		panic(err)
	}

	fmt.Println(completion.Choices[0].Message.Content)
}

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40import com.openai.client.OpenAIClient;
import com.openai.client.okhttp.OpenAIOkHttpClient;
import com.openai.models.chat.completions.ChatCompletionContentPart;
import com.openai.models.chat.completions.ChatCompletionContentPartText;
import com.openai.models.chat.completions.ChatCompletionCreateParams;
import com.openai.models.chat.completions.ChatCompletionUserMessageParam;
import java.io.IOException;
import java.nio.file.Files;
import java.nio.file.Path;
import java.util.Base64;
import java.util.List;

String encoded =
    Base64.getEncoder()
        .encodeToString(Files.readAllBytes(Path.of(System.getenv("OPENAI_EXAMPLE_FILE_PATH"))));
ChatCompletionContentPart file =
    ChatCompletionContentPart.ofFile(
        ChatCompletionContentPart.File.builder()
            .file(
                ChatCompletionContentPart.File.FileObject.builder()
                    .filename("document.pdf")
                    .fileData("data:application/pdf;base64," + encoded)
                    .build())
            .build());
ChatCompletionContentPart text =
    ChatCompletionContentPart.ofText(
        ChatCompletionContentPartText.builder().text("Summarize this document.").build());

ChatCompletionCreateParams params =
    ChatCompletionCreateParams.builder()
        .model("gpt-6-astra")
        .addMessage(
            ChatCompletionUserMessageParam.builder()
                .contentOfArrayOfContentParts(List.of(file, text))
                .build())
        .build();

client.chat().completions().create(params).choices().stream()
    .flatMap(choice -> choice.message().content().stream())
    .forEach(System.out::println);

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22using OpenAI.Chat;
#pragma warning disable OPENAI001

string key = Environment.GetEnvironmentVariable("OPENAI_API_KEY")!;
string model = "gpt-6-astra";
ChatClient client = new(model, key);

BinaryData fileBytes = BinaryData.FromBytes(await File.ReadAllBytesAsync("draconomicon.pdf"));
ChatCompletion completion = await client.CompleteChatAsync(
    new UserChatMessage(
        [
            ChatMessageContentPart.CreateFilePart(
                fileBytes,
                "application/pdf",
                "draconomicon.pdf"
            ),
            ChatMessageContentPart.CreateTextPart("What is the first dragon in the book?"),
        ]
    )
);

Console.WriteLine(completion.Content[0].Text);

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28require "base64"
require "openai"

client = OpenAI::Client.new
pdf_data = Base64.strict_encode64(File.binread("draconomicon.pdf"))
completion = client.chat.completions.create(
  model: "gpt-6-astra",
  messages: [
    {
      role: :user,
      content: [
        {
          type: :file,
          file: {
            filename: "document.pdf",
            file_data: "data:application/pdf;base64,#{pdf_data}"
          }
        },
        {
          type: :text,
          text: "Summarize this document."
        }
      ]
    }
  ]
)

puts(completion.choices.fetch(0).message.content)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24curl "https://api.openai.com/v1/chat/completions" \
    -H "Content-Type: application/json" \
    -H "Authorization: Bearer $OPENAI_API_KEY" \
    -d '{
        "model": "gpt-6-astra",
        "messages": [
            {
                "role": "user",
                "content": [
                    {
                        "type": "file",
                        "file": {
                            "filename": "draconomicon.pdf",
                            "file_data": "...base64 encoded bytes here..."
                        }
                    },
                    {
                        "type": "text",
                        "text": "What is the first dragon in the book?"
                    }
                ]
            }
        ]
    }'

使用注意事项

使用文件输入时，请注意以下限制：

Token 用量： PDF 解析会将提取的文本和页面图像都包含在上下文中，这可能会增加 Token 用量。在 Responses API 中，可将 detail 设为 auto（默认值）、low 或 high，以控制 PDF 页面图像的视觉细节量。在大规模部署之前，请了解定价和对 Token 用量的影响。了解更多定价信息。
文件大小限制： 单个请求可以包含多个文件，但每个文件必须小于 50 MB。请求中所有文件的总大小上限为 50 MB。
支持的模型： 同时包含文本和页面图像的 PDF 解析需要具有视觉能力的模型，例如 gpt-4o 及更新的模型。
文件上传用途： 您可以使用任何受支持的用途上传文件，但对于计划作为模型输入传入的文件，请使用 user_data。

支持的文件类型完整列表

此列表适用于 Responses API。Chat Completions 仅支持 .pdf （application/pdf），无论使用 file_data 还是 file_id。

类别	扩展名	MIME 类型
PDF 文件	PDF 文件（`.pdf`）	`application/pdf`
电子表格	Excel 工作表（`.xla`、`.xlb`、`.xlc`、`.xlm`、`.xls`、`.xlsx`、`.xlt`、`.xlw`）	`application/vnd.openxmlformats-officedocument.spreadsheetml.sheet`、`application/vnd.ms-excel`
电子表格	CSV / TSV / IIF（`.csv`、`.tsv`、`.iif`）、Google Sheets	`text/csv`、`application/csv`、`text/tsv`、`text/x-iif`、`application/x-iif`、`application/vnd.google-apps.spreadsheet`
富文本文档	Word/ODT/RTF 文档（`.doc`、`.docx`、`.dot`、`.odt`、`.rtf`）、Pages、Google Docs	`application/vnd.openxmlformats-officedocument.wordprocessingml.document`、`application/msword`、`application/rtf`、`text/rtf`、`application/vnd.oasis.opendocument.text`、`application/vnd.apple.pages`、`application/vnd.google-apps.document`、`application/vnd.apple.iwork`
演示文稿	PowerPoint 幻灯片（`.pot`、`.ppa`、`.pps`、`.ppt`、`.pptx`、`.pwz`、`.wiz`）、Keynote、Google Slides	`application/vnd.openxmlformats-officedocument.presentationml.presentation`、`application/vnd.ms-powerpoint`、`application/vnd.apple.keynote`、`application/vnd.google-apps.presentation`、`application/vnd.apple.iwork`
文本和代码	文本/代码格式（`.asm`、`.bat`、`.c`、`.cc`、`.conf`、`.cpp`、`.css`、`.cxx`、`.def`、`.dic`、`.eml`、`.h`、`.hh`、`.htm`、`.html`、`.ics`、`.ifb`、`.in`、`.js`、`.json`、`.ksh`、`.list`、`.log`、`.markdown`、`.md`、`.mht`、`.mhtml`、`.mime`、`.mjs`、`.nws`、`.pl`、`.py`、`.rst`、`.s`、`.sql`、`.srt`、`.text`、`.txt`、`.vcf`、`.vtt`、`.xml`）	`application/javascript`、`application/typescript`、`text/xml`、`text/x-shellscript`、`text/x-rst`、`text/x-makefile`、`text/x-lisp`、`text/x-asm`、`text/vbscript`、`text/css`、`message/rfc822`、`application/x-sql`、`application/x-scala`、`application/x-rust`、`application/x-powershell`、`text/x-diff`、`text/x-patch`、`application/x-patch`、`text/plain`、`text/markdown`、`text/x-java`、`text/x-script.python`、`text/x-python`、`text/x-c`、`text/x-c++`、`text/x-golang`、`text/html`、`text/x-php`、`application/x-php`、`application/x-httpd-php`、`application/x-httpd-php-source`、`text/x-ruby`、`text/x-sh`、`text/x-bash`、`application/x-bash`、`text/x-zsh`、`text/x-tex`、`text/x-csharp`、`application/json`、`text/x-typescript`、`text/javascript`、`text/x-go`、`text/x-rust`、`text/x-scala`、`text/x-kotlin`、`text/x-swift`、`text/x-lua`、`text/x-r`、`text/x-R`、`text/x-julia`、`text/x-perl`、`text/x-objectivec`、`text/x-objectivec++`、`text/x-erlang`、`text/x-elixir`、`text/x-haskell`、`text/x-clojure`、`text/x-groovy`、`text/x-dart`、`text/x-awk`、`application/x-awk`、`text/jsx`、`text/tsx`、`text/x-handlebars`、`text/x-mustache`、`text/x-ejs`、`text/x-jinja2`、`text/x-liquid`、`text/x-erb`、`text/x-twig`、`text/x-pug`、`text/x-jade`、`text/x-tmpl`、`text/x-cmake`、`text/x-dockerfile`、`text/x-gradle`、`text/x-ini`、`text/x-properties`、`text/x-protobuf`、`application/x-protobuf`、`text/x-sql`、`text/x-sass`、`text/x-scss`、`text/x-less`、`text/x-hcl`、`text/x-terraform`、`application/x-terraform`、`text/x-toml`、`application/x-toml`、`application/graphql`、`application/x-graphql`、`text/x-graphql`、`application/x-ndjson`、`application/json5`、`application/x-json5`、`text/x-yaml`、`application/toml`、`application/x-yaml`、`application/yaml`、`text/x-astro`、`text/srt`、`application/x-subrip`、`text/x-subrip`、`text/vtt`、`text/x-vcard`、`text/calendar`

后续步骤

接下来，您可以探索以下资源：

在 Playground 中试用文件输入

使用 Playground 编写并迭代包含文件输入的提示。

完整 API 参考

查看 API 参考，了解更多选项。

使用文件搜索检索大型语料库

如果您需要可扩展的搜索功能，请对分块后的文件进行检索，而不是在单个上下文窗口中发送完整文件。

使用托管式 Shell 深入分析电子表格

使用托管式 Shell 处理连接、聚合和图表绘制等高级电子表格工作流。

建议

开始使用

核心概念

SDK 与 CLI

资源

旧版 API

选择模型

文本与代码

提示词

推理

图像与视频

实时交互与音频

专用模型

智能体 API

智能体 SDK

ChatKit

搜索与检索

连接工具与数据

构建工具工作流

计算机与代码

媒体

GPT-Live

Realtime API

构建语音应用

连接

音频处理

上线

性能与质量

成本与吞吐量

安全与治理

基础设施与访问

运营

核心概念

Plan

Build

Test and publish

Conversion specs

指南

资源

开始使用

指南

File Upload

API

Measurement

Advertiser API

API 参考

开始使用

基础

探索

可用平台

发布

工作流

能力

参考资料

自定义

配置文件

智能体配置

扩展 ChatGPT 和 Codex 的功能

Linux

Windows

开发工作流

扩展并实现自动化

环境

使用 Codex 构建

第三方集成

参考资料

权限

Codex Security

网络安全

入门

ChatGPT Work

身份与身份验证

工作空间访问权限、策略和模型

插件和连接器控制

使用情况、治理与合规

部署和模型提供商

社区

博客

社区

博客