fal-ai/bytedance/seed/v2/mini

Seed 2.0 Mini is a high-performance multimodal model optimized for low latency and high concurrency. It supports text, image, and video input with 256K context and configurable thinking/reasoning modes.

Inference

Commercial use

Schema

LLMs

Playground API

Input

Prompt*

Image Urls

Hint: Drag and drop files from your computer, images from web pages, paste from clipboard (Ctrl/Cmd+V), or provide a URL.

Video Urls

Additional Settings

Customize your input with more control.

Result

Idle

{
  "output": "",
  "messages": [
    {
      "role": "system"
    }
  ]
}

What would you like to do next?

Your request will cost $0.0001 per 1000 units. For inputs under 128k tokens, the units per input token is 1. For inputs of over 128k tokens, 2 units will be charged per token. Similarly, each output token costs 4 units, provided the total output length (reasoning + output) is under 128k tokens, and 8 units per token otherwise.

fal-ai/bytedance/seed/v2/mini

Input

Result

What would you like to do next?

Logs