Interactive Playground
Test prompts in real-time with variable inputs, model configuration, and streaming responses.
What is the Playground?#
The Playground is an interactive environment for testing and iterating on your prompts. It provides real-time feedback, streaming responses, and the ability to quickly experiment with different inputs and model configurations.
Real-time Streaming
See responses as they're generated with streaming output.
Variable Testing
Fill in variables and test different input combinations.
Model Configuration
Adjust temperature, max tokens, and other model parameters.
Execution History
Review past runs and compare outputs across sessions.
Running Prompts#
To run a prompt in the Playground:
Open a Prompt
Fill Variables
Configure Model
Click Run
Keyboard Shortcut
Ctrl+Enter (or Cmd+Enter on Mac) to quickly run your prompt without clicking the button.Model Configuration#
The Playground allows you to configure various model parameters:
| Parameter | Description | Range |
|---|---|---|
| Model | The AI model to use (GPT-4, Claude, etc.) | Varies by provider |
| Temperature | Controls randomness in responses | 0.0 - 2.0 |
| Max Tokens | Maximum length of the response | 1 - model limit |
| Top P | Nucleus sampling threshold | 0.0 - 1.0 |
| Frequency Penalty | Reduces repetition of tokens | -2.0 - 2.0 |
| Presence Penalty | Encourages topic diversity | -2.0 - 2.0 |
Model Selection
Working with Variables#
Variables in your prompt are automatically detected and displayed in the Playground. You can fill them in manually or load from saved presets.
You are a {{role}} helping a customer with {{issue_type}}.
Customer message: {{customer_message}}
Respond professionally and helpfully.The Playground will show input fields for role, issue_type, and customer_message. You can save commonly used combinations as presets for quick access.
Execution History#
Every execution in the Playground is automatically saved to your history. This allows you to:
- Review past outputs and compare different runs
- Restore previous variable values
- Track token usage and response times
- Export successful runs as test cases
Response Time
See how long each execution took.
Token Usage
Track input and output token counts.
Quality Scores
View auto-generated quality metrics.
Tips & Shortcuts#
Make the most of the Playground with these tips:
| Shortcut | Action |
|---|---|
| Ctrl+Enter | Run the current prompt |
| Ctrl+S | Save current prompt as new version |
| Ctrl+Shift+P | Open preset selector |
| Esc | Cancel running execution |
| Ctrl+L | Clear output |