LLM Evals
Playground
Chat with any configured model directly to test prompts before running experiments.
The playground is listed in the sidebar but not yet available. This page will be enabled in a future release.
What the playground will do
The playground will be a chat interface for talking to any model your organization has configured. You'll be able to test prompts, compare model responses and verify API key connectivity without setting up a formal experiment.
Planned features
- Pick a provider and model from a dropdown, then chat in real time
- Multi-turn conversations with full history sent on each request
- Switch models mid-session (clears conversation)
- No project selection required, available from the sidebar at any time
You'll need at least one API key configured in Settings before the playground can connect to a model.
PreviousCI/CD integration
NextLeaderboard