Multimodal

Send text, image, audio, or video inputs when the selected model supports them.

Content parts

OpenAI-compatible chat requests can use an array of content parts for multimodal models, such as text, image_url, input_audio, or provider-specific file/video parts. Native Claude and Gemini requests use their own content schemas. Do not assume every model accepts every media type.

Example image input

{
  "model": "google/gemini-3.1-flash-lite",
  "messages": [{
    "role": "user",
    "content": [
      {"type": "text", "text": "Describe this image."},
      {"type": "image_url", "image_url": {"url": "https://example.com/image.png"}}
    ]
  }]
}

Base64 and local files

No URL? Inline the image as a base64 data URI in the same image_url field. This avoids hosting the file but increases request size, so prefer URLs for large or reused images.

{
  "type": "image_url",
  "image_url": { "url": "data:image/png;base64,iVBORw0KGgoAAAANSUh..." }
}

Audio input

Audio-capable chat models accept an input_audio part with base64-encoded audio and its format. Use this to ask questions about a recording or transcribe inline.

{
  "role": "user",
  "content": [
    {"type": "text", "text": "What is said in this clip?"},
    {"type": "input_audio", "input_audio": {"data": "<base64-audio>", "format": "mp3"}}
  ]
}

Model-dependent features

OneRouter keeps protocol compatibility and forwards model-specific parameters where supported. Features such as tool calling, structured output, reasoning, prompt caching, web search, image input, audio input, video generation, and provider-specific options depend on the selected model and upstream provider. Check the model detail page before relying on a feature in production.