Skip to main content

Documentation Index

Fetch the complete documentation index at: https://runcrate.ai/docs/llms.txt

Use this file to discover all available pages before exploring further.

Every model on Runcrate has an interactive playground in the dashboard. Test models, experiment with parameters, and get auto-generated code — without writing any code yourself.

​
Accessing the Playground

  1. Go to Dashboard → Models
  2. Click on any model
  3. The playground opens with the model ready to use

​
Playground Types

The playground adapts to the model’s category:

​
Chat Playground

For Chat, Code, Reasoning, and Vision models.
  • Real-time streaming — Responses appear token-by-token
  • System prompt — Set the model’s behavior
  • Multi-turn conversation — Full conversation history
  • Temperature and max tokens — Adjustable via settings panel
  • Image uploads — For vision-capable models, attach images to your messages
  • Reasoning blocks — For reasoning models, thinking steps are shown in collapsible blocks

​
Image Playground

For image generation models.
  • Prompt input — Describe the image you want
  • Parameter controls — Aspect ratio, steps, guidance, seed, and model-specific settings
  • Live preview — Generated images display inline
  • Reference images — Upload images for img2img, editing, and Kontext models

​
Video Playground

For video generation models.
  • Prompt input — Describe the video
  • Duration selector — Choose from available durations (buttons or slider depending on model)
  • Async status — Shows job status (queued → processing → completed)
  • Video player — Watch the result directly in the browser
  • Download — Save the MP4

​
TTS Playground

For text-to-speech models.
  • Text input — Enter the text to speak
  • Voice selector — Choose from available voice presets
  • Audio player — Listen to the result with streaming playback

​
ASR Playground

For speech-to-text models.
  • File upload — Upload an audio file (MP3, WAV, etc.)
  • Transcription output — View the transcribed text
  • Duration and language — Shows detected language and audio duration

​
Auto-Generated Code

Every playground shows a code panel with working examples in three languages:
  • curl — Direct HTTP request
  • Python — Using the OpenAI SDK
  • JavaScript — Using the OpenAI SDK
The code updates automatically based on your selected model and parameters. Copy it directly into your application.

​
API Key

The playground uses your workspace’s default API key. If you don’t have one yet, it’s created automatically when you first open a model.