For AI agents: a documentation index is available at the root level at /llms.txt and /llms-full.txt. Append /llms.txt to any URL for a page-level index, or .md for the markdown version of any page.
Copy to LLMGithubGo to App
DocumentationIntegrationsAgent OptimizationSelf-hosting OpikSDK & API referenceOpik University
DocumentationIntegrationsAgent OptimizationSelf-hosting OpikSDK & API referenceOpik University
    • Overview
  • Intro
    • Opik Overview
    • Next steps / Set expectations
  • Observability
    • Log Traces
    • Annotate Traces
  • Evaluation
    • Evaluation Concepts and Overview
    • Create Evaluation Datasets
    • Define Evaluation Metrics
    • Evaluate your LLM Application
    • No-code LLM Evaluation Workflow
  • Prompt Engineering
    • Prompt Management
    • Prompt Playground
  • Testing
    • PyTest Integration
  • Production Monitoring
    • Online Evaluation Rules
LogoLogo
Copy to LLMGithubGo to App
On this page
  • Interactive Prompt Engineering with Opik’s Playground
  • Key Highlights
Prompt Engineering

Prompt Playground

Was this page helpful?
Previous

PyTest Integration

Next
Built with

Interactive Prompt Engineering with Opik’s Playground

This video demonstrates how to use Opik’s prompting playground for hands-on prompt engineering and testing. The playground provides an interactive environment where you can craft prompts, test variations, and run experiments directly within the UI. You’ll learn how to leverage versioned prompts from your library, conduct A/B testing, and seamlessly transition from prompt development to full dataset evaluation.

Key Highlights

  • Parallel A/B Testing: Run multiple prompt variations side-by-side to compare outputs and identify the most effective approaches for your use case
  • Flexible Prompt Types: Switch between system and user prompts depending on your workflow needs, from simple queries to complex system instructions
  • Versioned Prompt Integration: Load prompts directly from your managed prompt library, always accessing the latest committed version for consistent testing
  • Dataset-Driven Experiments: Integrate test datasets directly into playground sessions to run comprehensive evaluations right from the interface
  • Configurable Model Parameters: Adjust creativity levels, output token limits, and other model settings to fine-tune behavior for your specific requirements
  • Template Variable Support: Use mustache/handlebar notation (e.g., {{question}}) to create dynamic prompts that work with dataset variables
  • Live Version Control: Save prompt modifications directly from the playground with commit messages and metadata, creating new versions in your prompt library
  • Instant Experiment Creation: Generate experiments with visual feedback banners, then view comprehensive results including outputs from your Q&A datasets
  • Seamless Workflow Integration: Bridge the gap between prompt development, testing, and full evaluation without leaving the Opik environment