Skip to main contentThe Playground is an interactive space where you can experiment with your prompts, test different LLMs, adjust settings, and instantly see the results to craft the perfect prompt for your tasks.
Features
Running Prompts
Tool Calls
In the Playground you can run prompts with different possibilities, which include:
- Run your prompts with production test cases, switch between top LLMs and their configurations to assess outputs.
- Access and review previous test sessions, thanks to a versioned history of past runs with the ability to rollback to any version.
- Pick up from any point in your conversation history, and build on previous outputs.
If your prompt needs to rely on external services, the Playground gets you covered as it supports complete tool calling, which includes:
- Manually adding tool responses and getting the final LLM outputs.
- Manage and modify complex conversation threads that include tool calling or connecting to MCP servers to drive the model’s response.
