> ## Documentation Index
> Fetch the complete documentation index at: https://thethirdpenco-feat-tool-lifecycle-events.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Multimodal Chatbot

> Handle images and files in conversations

Build a chatbot that can analyze images, PDFs, and other files alongside text.

## Basic Multimodal Chatbot

Messages can contain multiple parts (text, images, files).

```python theme={null}
import asyncio
import base64
from ai_query import generate_text
from ai_query.providers import google

class MultimodalChatbot:
    def __init__(self, model, system_prompt: str = None):
        self.model = model
        self.system_prompt = system_prompt
        self.messages: list[dict] = []

    async def chat(
        self,
        text: str,
        image_path: str | None = None,
        image_url: str | None = None
    ) -> str:
        # Build content list
        content = [{"type": "text", "text": text}]

        if image_path:
            with open(image_path, "rb") as f:
                image_data = base64.b64encode(f.read()).decode()
            content.append({
                "type": "image",
                "image": image_data,
                "media_type": "image/png"
            })
        elif image_url:
            content.append({
                "type": "image",
                "image": image_url
            })

        self.messages.append({"role": "user", "content": content})

        result = await generate_text(
            model=self.model,
            system=self.system_prompt,
            messages=self.messages
        )

        self.messages.append({"role": "assistant", "content": result.text})
        return result.text

async def main():
    bot = MultimodalChatbot(
        model=google("gemini-2.0-flash"),
        system_prompt="You are a helpful assistant that can analyze images."
    )

    # Analyze an image file
    response = await bot.chat("What's in this image?", image_path="photo.jpg")
    print(response)

    # Follow-up question (remembers the image)
    response = await bot.chat("What colors do you see?")
    print(response)

    # Analyze image from URL
    response = await bot.chat(
        "Describe this image.",
        image_url="https://example.com/image.jpg"
    )
    print(response)

asyncio.run(main())
```

## Using Message Objects

You can also use `Message` and content part objects for type safety:

```python theme={null}
from ai_query import generate_text
from ai_query.providers import google
from ai_query.types import Message, TextPart, ImagePart

messages = [
    Message(
        role="user",
        content=[
            TextPart(text="What is this?"),
            ImagePart(image="https://example.com/image.jpg")
        ]
    )
]

result = await generate_text(model=google("gemini-2.0-flash"), messages=messages)
```

## PDF Analysis Chatbot

```python theme={null}
import asyncio
import base64
from ai_query import generate_text
from ai_query.providers import google

class PDFChatbot:
    def __init__(self, model):
        self.model = model
        self.messages: list[dict] = []

    async def load_pdf(self, pdf_path: str, question: str) -> str:
        """Load a PDF and ask a question about it."""
        with open(pdf_path, "rb") as f:
            pdf_data = base64.b64encode(f.read()).decode()

        content = [
            {"type": "text", "text": question},
            {"type": "file", "data": pdf_data, "media_type": "application/pdf"}
        ]

        self.messages.append({"role": "user", "content": content})

        result = await generate_text(
            model=self.model,
            messages=self.messages
        )

        self.messages.append({"role": "assistant", "content": result.text})
        return result.text

    async def ask(self, question: str) -> str:
        """Ask follow-up questions about the loaded document."""
        self.messages.append({"role": "user", "content": question})

        result = await generate_text(
            model=self.model,
            messages=self.messages
        )

        self.messages.append({"role": "assistant", "content": result.text})
        return result.text

async def main():
    bot = PDFChatbot(google("gemini-2.0-pro"))

    # Load PDF and ask initial question
    response = await bot.load_pdf("report.pdf", "Summarize the key findings.")
    print(response)

    # Follow-up questions
    response = await bot.ask("What were the main conclusions?")
    print(response)

    response = await bot.ask("Any recommendations mentioned?")
    print(response)

asyncio.run(main())
```

## Multi-Image Comparison

```python theme={null}
import asyncio
import base64
from ai_query import generate_text
from ai_query.providers import anthropic

async def compare_images(image_paths: list[str], question: str) -> str:
    """Compare multiple images."""
    content = [{"type": "text", "text": question}]

    for path in image_paths:
        with open(path, "rb") as f:
            image_data = base64.b64encode(f.read()).decode()
        content.append({
            "type": "image",
            "image": image_data,
            "media_type": "image/png"
        })

    result = await generate_text(
        model=anthropic("claude-3-5-sonnet-20241022"),
        messages=[{"role": "user", "content": content}]
    )

    return result.text

async def main():
    response = await compare_images(
        ["before.png", "after.png"],
        "Compare these two images. What changed?"
    )
    print(response)

asyncio.run(main())
```

## Screenshot Debug Helper

```python theme={null}
import asyncio
import base64
from ai_query import generate_text
from ai_query.providers import google

async def debug_screenshot(
    screenshot_path: str,
    error_log: str | None = None
) -> str:
    """Analyze a screenshot and optional error log for debugging."""
    content = [{"type": "text", "text": "Help me debug this issue."}]

    # Add screenshot
    with open(screenshot_path, "rb") as f:
        screenshot_data = base64.b64encode(f.read()).decode()
    content.append({
        "type": "image",
        "image": screenshot_data,
        "media_type": "image/png"
    })

    # Add error log if provided
    if error_log:
        content.append({
            "type": "file",
            "data": error_log,
            "media_type": "text/plain"
        })

    result = await generate_text(
        model=google("gemini-2.0-pro"),
        system="You are a debugging expert. Analyze the screenshot and logs to identify issues.",
        messages=[{"role": "user", "content": content}]
    )

    return result.text

async def main():
    # Read error log
    with open("error.log") as f:
        error_log = f.read()

    response = await debug_screenshot("error_screenshot.png", error_log)
    print(response)

asyncio.run(main())
```