Thousands of engineers use the LangSmith Playground as their IDE for prompt engineering. Now, we’ve made the Playground even better — here’s what’s new:

• Expanded model support - Experiment with the latest models, like Llama 3.1, Mistral Large, Gemma 2, and Gemini 1.5 Pro, with increased access to various providers, including Vertex AI, Mistral, and Google Generative AI

• Concurrent runs in one click - Instantly test output consistency by simultaneously generating up to five outputs from same prompt in a single click

• Enhanced tool compatibility - Tool use is supported in the Playground for all LangChain models that support tool calling - giving you even more flexibility in your testing and development

• Unified layout - The playground is deeply embedded within LangSmith. Access the same standardized interface whether you’re debugging a trace, creating a new prompt, or evaluating across a dataset

Explore our upgraded playground in LangSmith: http://smith.langchain.com/playground