Vision and image input

Send images to multimodal models — encoding, size limits, and what it costs.

Claude and Gemini models accept images alongside text. Images are content blocks inside an ordinary message.

Base64 image

Copy{
  "model": "claude-sonnet-4-6",
  "max_tokens": 1024,
  "messages": [{
    "role": "user",
    "content": [
      {"type": "image", "source": {
        "type": "base64",
        "media_type": "image/png",
        "data": "iVBORw0KGgoAAAANSUhEUg..."
      }},
      {"type": "text", "text": "What does this chart show?"}
    ]
  }]
}

Python

Copyimport base64, anthropic

img = base64.standard_b64encode(open(class="s">"chart.png", class="s">"rb").read()).decode()

client = anthropic.Anthropic(api_key=class="s">"sk-your-key", base_url=class="s">"https://aiprimetech.io")
msg = client.messages.create(
    model=class="s">"claude-sonnet-4-6",
    max_tokens=1024,
    messages=[{class="s">"role": class="s">"user", class="s">"content": [
        {class="s">"type": class="s">"image", class="s">"source": {class="s">"type": class="s">"base64",
         class="s">"media_type": class="s">"image/png", class="s">"data": img}},
        {class="s">"type": class="s">"text", class="s">"text": class="s">"What does this chart show?"},
    ]}],
)
print(msg.content[0].text)

Practical limits