Skip to content
Docs

AI Gateway File and PDF Input

Send a document with a question to summarize its contents or extract information. These examples send a PDF directly in the request. For photos and screenshots, see vision.

Before you start

Set AI_GATEWAY_API_KEY, install your SDK or API client, and save a PDF as document.pdf in your working directory. Use a model that supports PDF input.

File support depends on the model and provider. A file part does not make every document format compatible. Extract text from unsupported office documents before sending it as a text part. For scanned pages, choose a model that can also interpret the page images.

Summarize a PDF

These examples send the PDF bytes and request a summary. The SDK examples print the summary; cURL prints the JSON response containing it:

file-input.ts
import { readFileSync } from 'node:fs';
import { generateText } from 'ai';
 
const { text } = await generateText({
  model: 'anthropic/claude-sonnet-5',
  messages: [
    {
      role: 'user',
      content: [
        { type: 'text', text: 'Summarize this document in one sentence.' },
        {
          type: 'file',
          data: readFileSync('document.pdf'),
          mediaType: 'application/pdf',
        },
      ],
    },
  ],
});
 
console.log(text);
file-input_ai.py
import asyncio
from pathlib import Path
import ai
 
 
async def main():
    model = ai.get_model("anthropic/claude-sonnet-5")
    messages = [
        ai.user_message(
            "Summarize this document in one sentence.",
            ai.file_part(
                Path("document.pdf").read_bytes(), media_type="application/pdf"
            ),
        )
    ]
    async with ai.stream(model, messages) as stream:
        async for event in stream:
            if isinstance(event, ai.events.TextDelta):
                print(event.chunk, end="", flush=True)
    print()
 
 
asyncio.run(main())
file-input-chat.ts
import { readFileSync } from 'node:fs';
import OpenAI from 'openai';
 
const client = new OpenAI({
  apiKey: process.env.AI_GATEWAY_API_KEY,
  baseURL: 'https://ai-gateway.vercel.sh/v1',
});
const data = readFileSync('document.pdf').toString('base64');
 
const response = await client.chat.completions.create({
  model: 'anthropic/claude-sonnet-5',
  messages: [
    {
      role: 'user',
      content: [
        {
          type: 'text',
          text: 'Summarize this document in one sentence.',
        },
        {
          type: 'file',
          file: {
            filename: 'document.pdf',
            file_data: `data:application/pdf;base64,${data}`,
          },
        },
      ],
    },
  ],
});
 
console.log(response.choices[0]?.message.content);
file-input_chat.py
import base64
import os
from pathlib import Path
from openai import OpenAI
 
client = OpenAI(
    api_key=os.environ["AI_GATEWAY_API_KEY"],
    base_url="https://ai-gateway.vercel.sh/v1",
)
data = base64.b64encode(Path("document.pdf").read_bytes()).decode("ascii")
 
response = client.chat.completions.create(
    model="anthropic/claude-sonnet-5",
    messages=[
        {
            "role": "user",
            "content": [
                {"type": "text", "text": "Summarize this document in one sentence."},
                {
                    "type": "file",
                    "file": {
                        "filename": "document.pdf",
                        "file_data": f"data:application/pdf;base64,{data}",
                    },
                },
            ],
        }
    ],
)
 
print(response.choices[0].message.content)
file-input-chat.sh
FILE_BASE64=$(base64 < document.pdf | tr -d '\n')
 
printf '{
  "model": "anthropic/claude-sonnet-5",
  "messages": [
    {
      "role": "user",
      "content": [
        {
          "type": "text",
          "text": "Summarize this document in one sentence."
        },
        {
          "type": "file",
          "file": {
            "filename": "document.pdf",
            "file_data": "data:application/pdf;base64,%s"
          }
        }
      ]
    }
  ]
}' "$FILE_BASE64" | \
curl --fail-with-body https://ai-gateway.vercel.sh/v1/chat/completions \
  -H "Authorization: Bearer $AI_GATEWAY_API_KEY" \
  -H "Content-Type: application/json" \
  --data-binary @-
file-input-messages.ts
import { readFileSync } from 'node:fs';
import Anthropic from '@anthropic-ai/sdk';
 
const client = new Anthropic({
  apiKey: process.env.AI_GATEWAY_API_KEY,
  baseURL: 'https://ai-gateway.vercel.sh',
});
const data = readFileSync('document.pdf').toString('base64');
 
const response = await client.messages.create({
  model: 'anthropic/claude-sonnet-5',
  max_tokens: 1024,
  messages: [
    {
      role: 'user',
      content: [
        {
          type: 'text',
          text: 'Summarize this document in one sentence.',
        },
        {
          type: 'document',
          source: {
            type: 'base64',
            media_type: 'application/pdf',
            data: data,
          },
        },
      ],
    },
  ],
});
 
for (const block of response.content) {
  if (block.type === 'text') console.log(block.text);
}
file-input_messages.py
import base64
import os
from pathlib import Path
from anthropic import Anthropic
 
client = Anthropic(
    api_key=os.environ["AI_GATEWAY_API_KEY"],
    base_url="https://ai-gateway.vercel.sh",
)
data = base64.b64encode(Path("document.pdf").read_bytes()).decode("ascii")
 
response = client.messages.create(
    model="anthropic/claude-sonnet-5",
    max_tokens=1024,
    messages=[
        {
            "role": "user",
            "content": [
                {"type": "text", "text": "Summarize this document in one sentence."},
                {
                    "type": "document",
                    "source": {
                        "type": "base64",
                        "media_type": "application/pdf",
                        "data": data,
                    },
                },
            ],
        }
    ],
)
 
for block in response.content:
    if block.type == "text":
        print(block.text)
file-input-messages.sh
FILE_BASE64=$(base64 < document.pdf | tr -d '\n')
 
printf '{
  "model": "anthropic/claude-sonnet-5",
  "max_tokens": 1024,
  "messages": [
    {
      "role": "user",
      "content": [
        {
          "type": "text",
          "text": "Summarize this document in one sentence."
        },
        {
          "type": "document",
          "source": {
            "type": "base64",
            "media_type": "application/pdf",
            "data": "%s"
          }
        }
      ]
    }
  ]
}' "$FILE_BASE64" | \
curl --fail-with-body https://ai-gateway.vercel.sh/v1/messages \
  -H "Authorization: Bearer $AI_GATEWAY_API_KEY" \
  -H "Content-Type: application/json" \
  -H "anthropic-version: 2023-06-01" \
  --data-binary @-
file-input-responses.ts
import { readFileSync } from 'node:fs';
import OpenAI from 'openai';
 
const client = new OpenAI({
  apiKey: process.env.AI_GATEWAY_API_KEY,
  baseURL: 'https://ai-gateway.vercel.sh/v1',
});
const data = readFileSync('document.pdf').toString('base64');
 
const response = await client.responses.create({
  model: 'anthropic/claude-sonnet-5',
  input: [
    {
      role: 'user',
      content: [
        {
          type: 'input_text',
          text: 'Summarize this document in one sentence.',
        },
        {
          type: 'input_file',
          filename: 'document.pdf',
          file_data: `data:application/pdf;base64,${data}`,
        },
      ],
    },
  ],
});
 
console.log(response.output_text);
file-input_responses.py
import base64
import os
from pathlib import Path
from openai import OpenAI
 
client = OpenAI(
    api_key=os.environ["AI_GATEWAY_API_KEY"],
    base_url="https://ai-gateway.vercel.sh/v1",
)
data = base64.b64encode(Path("document.pdf").read_bytes()).decode("ascii")
 
response = client.responses.create(
    model="anthropic/claude-sonnet-5",
    input=[
        {
            "role": "user",
            "content": [
                {
                    "type": "input_text",
                    "text": "Summarize this document in one sentence.",
                },
                {
                    "type": "input_file",
                    "filename": "document.pdf",
                    "file_data": f"data:application/pdf;base64,{data}",
                },
            ],
        }
    ],
)
 
print(response.output_text)
file-input-responses.sh
FILE_BASE64=$(base64 < document.pdf | tr -d '\n')
 
printf '{
  "model": "anthropic/claude-sonnet-5",
  "input": [
    {
      "role": "user",
      "content": [
        {
          "type": "input_text",
          "text": "Summarize this document in one sentence."
        },
        {
          "type": "input_file",
          "filename": "document.pdf",
          "file_data": "data:application/pdf;base64,%s"
        }
      ]
    }
  ]
}' "$FILE_BASE64" | \
curl --fail-with-body https://ai-gateway.vercel.sh/v1/responses \
  -H "Authorization: Bearer $AI_GATEWAY_API_KEY" \
  -H "Content-Type: application/json" \
  --data-binary @-

API format differences

See the AI SDK file-part guide and Python beta multimodal-message guide for constructing SDK messages. The table below compares those messages with the HTTP formats.

FormatInput representationDetails
AI SDK 7file part with mediaType: 'application/pdf'Pass bytes, a data URL, or a supported URL. See file parts.
AI SDK for Python (beta)ai.file_part(..., media_type="application/pdf")See Python file parts.
Chat Completionsfile with filename and a file_data data URLAI Gateway maps the file to the selected provider. See file requests.
Messages APIdocument with a base64 PDF sourceSee Messages documents.
Responses / OpenResponsesinput_file with filename and a file_data data URLFor non-OpenAI models, AI Gateway translates the request through OpenResponses. Native Responses models have their own input support.

Handling larger inputs

Inline base64 increases the request size. Check the selected model's context and file-size limits, and keep any fallback models compatible with the input type. When using a hosted URL, make sure the provider can retrieve it without your application's browser session.

For request details and failures, inspect AI Gateway logs. See Inputs & Tools to choose another capability.

Last updated September 8, 2026
Previous
Vision
Next
Tool Use

Was this helpful?

supported.