> ## Documentation Index
> Fetch the complete documentation index at: https://docs.opper.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Vision & PDFs

> Send images and PDF documents to the model as part of a chat message.

Put an image or a PDF into a chat message and the model can read it. It's the same `messages` array, with a richer `content` field — no separate endpoint. Vision and PDF are model *capabilities*, so you send the media to a regular chat model that supports them.

Not every model does. Filter the [catalog](/capabilities/models) by capability — or call `GET /v3/models?capability=vision` (images) or `?capability=pdf` (documents) — to find ones that accept each. The Claude, Gemini, and GPT families support both.

## Images

Two ways to send an image: a hosted URL or inline base64.

<CodeGroup>
  ```python Python theme={null}
  import os, base64
  from openai import OpenAI

  client = OpenAI(
      base_url="https://api.opper.ai/v3/compat",
      api_key=os.environ["OPPER_API_KEY"],
  )

  # Hosted URL
  r = client.chat.completions.create(
      model="openai/gpt-5-mini",
      messages=[{
          "role": "user",
          "content": [
              {"type": "text", "text": "What's in this picture?"},
              {"type": "image_url", "image_url": {"url": "https://example.com/cat.jpg"}},
          ],
      }],
  )
  print(r.choices[0].message.content)

  # Inline base64
  with open("cat.jpg", "rb") as f:
      b64 = base64.b64encode(f.read()).decode()
  r = client.chat.completions.create(
      model="openai/gpt-5-mini",
      messages=[{
          "role": "user",
          "content": [
              {"type": "text", "text": "What's in this picture?"},
              {"type": "image_url", "image_url": {"url": f"data:image/jpeg;base64,{b64}"}},
          ],
      }],
  )
  ```

  ```typescript TypeScript theme={null}
  import OpenAI from "openai";
  import { readFileSync } from "node:fs";

  const client = new OpenAI({
      baseURL: "https://api.opper.ai/v3/compat",
      apiKey: process.env.OPPER_API_KEY!,
  });

  // Hosted URL
  let r = await client.chat.completions.create({
      model: "openai/gpt-5-mini",
      messages: [{
          role: "user",
          content: [
              { type: "text", text: "What's in this picture?" },
              { type: "image_url", image_url: { url: "https://example.com/cat.jpg" } },
          ],
      }],
  });
  console.log(r.choices[0].message.content);

  // Inline base64
  const b64 = readFileSync("cat.jpg").toString("base64");
  r = await client.chat.completions.create({
      model: "openai/gpt-5-mini",
      messages: [{
          role: "user",
          content: [
              { type: "text", text: "What's in this picture?" },
              { type: "image_url", image_url: { url: `data:image/jpeg;base64,${b64}` } },
          ],
      }],
  });
  ```
</CodeGroup>

Use the URL form when the image is already on the web. Use base64 for local files or anything not publicly reachable.

## PDFs

PDFs work the same way. The model reads both the text and any embedded images (charts, diagrams, scanned pages).

```python Python theme={null}
import base64

with open("contract.pdf", "rb") as f:
    b64 = base64.b64encode(f.read()).decode()

r = client.chat.completions.create(
    model="openai/gpt-5-mini",
    messages=[{
        "role": "user",
        "content": [
            {"type": "text", "text": "Summarize the key clauses in this contract."},
            {"type": "file", "file": {"file_data": f"data:application/pdf;base64,{b64}"}},
        ],
    }],
)
print(r.choices[0].message.content)
```

## Free text or structured output

| Need                                                                           | Reach for                                                                    |
| ------------------------------------------------------------------------------ | ---------------------------------------------------------------------------- |
| Show an image and ask a free-text question about it                            | A plain message (this page)                                                  |
| Extract structured fields from an image or PDF (a receipt, an invoice, a form) | [Structured output](/build/gateway/structured-output) with `response_format` |
| Run a multi-turn conversation about an uploaded document                       | A plain message (this page)                                                  |
| Batch process documents into a database                                        | [Structured output](/build/gateway/structured-output)                        |

Add `response_format` when you want typed JSON out of an image or PDF. Leave it off when the model just needs to talk about the media.

## What's next

<CardGroup cols={2}>
  <Card title="Structured output" icon="braces" href="/build/gateway/structured-output">
    Multimodal input with typed JSON output.
  </Card>

  <Card title="Conversations" icon="comments" href="/build/gateway/conversations">
    Multi-turn chat. Works with image and PDF messages too.
  </Card>

  <Card title="Models" icon="brain" href="https://opper.ai/models">
    Which models accept which input types.
  </Card>
</CardGroup>