Multimodal

Images, documents, audio, and video are all available through the gateway with the same key and credits. Request and response shapes are the standard OpenAI-compatible ones; this page shows where each modality lives.

Need	Endpoint	Reference
Image or file input to a chat model	`POST /v1/chat/completions` with content parts	below
Parse an uploaded document	`plugins: [{ "id": "file-parser" }]`	below
Generate an image	`POST /v1/images`	Images and video
Generate a video	`POST /v1/videos`, then poll	Images and video
Text to speech	`POST /v1/audio/speech`	Audio
Transcribe audio	`POST /v1/audio/transcriptions`	Audio
Upload and reuse a file	`POST /v1/files`	Files

Image input

Send an image_url content part. Models that price image inputs show credits_per_image in the catalog.

curl --fail-with-body https://api.pre.dev/v1/chat/completions \
  -H "Authorization: Bearer $PREDEV_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "google/gemini-3.8-flash",
    "messages": [{
      "role": "user",
      "content": [
        { "type": "text", "text": "What is in this picture?" },
        { "type": "image_url", "image_url": { "url": "https://upload.wikimedia.org/wikipedia/commons/3/3a/Cat03.jpg" } }
      ]
    }]
  }'

const completion = await client.chat.completions.create({
  model: 'google/gemini-3.8-flash',
  messages: [{
    role: 'user',
    content: [
      { type: 'text', text: 'What is in this picture?' },
      { type: 'image_url', image_url: { url: 'https://upload.wikimedia.org/wikipedia/commons/3/3a/Cat03.jpg' } },
    ],
  }],
});

completion = client.chat.completions.create(
    model="google/gemini-3.8-flash",
    messages=[{
        "role": "user",
        "content": [
            {"type": "text", "text": "What is in this picture?"},
            {"type": "image_url", "image_url": {"url": "https://upload.wikimedia.org/wikipedia/commons/3/3a/Cat03.jpg"}},
        ],
    }],
)

Documents

Attach a PDF or other document as a file content part and add the file-parser plugin so any model can read it:

{
  "model": "anthropic/claude-sonnet-5",
  "plugins": [{ "id": "file-parser" }],
  "messages": [{
    "role": "user",
    "content": [
      { "type": "text", "text": "Summarize this contract in five bullets." },
      { "type": "file", "file": { "filename": "contract.pdf", "file_data": "data:application/pdf;base64,JVBERi0xLjQK..." } }
    ]
  }]
}

Upload once with POST /v1/files when the same document is used across many calls. Files are scoped to the workspace that uploaded them.

Web search

Add plugins: [{ "id": "web" }], or use a model’s :online variant, and tune with web_search_options. Search cost is part of usage.cost. See routing and fallbacks.

Generated media

Image, video, and speech models are listed by /v1/images/models, /v1/videos/models, and the main catalog. Video is asynchronous: submit, poll GET /v1/videos/{jobId}, then download from /content; the job is charged when a poll reports completion. See images and video and audio.

Get started

Guides

API reference

Image input

Documents

Web search

Generated media

​Image input

​Documents

​Web search

​Generated media

Image input

Documents

Web search

Generated media