Improved dataset creation: faster and more flexible
Uploading data from an existing CSV/JSONL files is one of the easiest ways to get started with datasets and experiments. We’ve revamped the upload flow to...
Product naming changes: LangSmith Deployment and LangSmith Studio
We’ve updated product names across the LangSmith UI, docs, and website to make it clearer how our tools fit together. What changed You’ll see these new names...
JavaScript Code Evals for faster eval workflows
JavaScript Code Evals allow you to write custom Code Evals in JavaScript in addition to Python. This gives you the flexibility to choose the language that...
Composite Evaluators in LangSmith
If you’re running multiple evaluators to measure different aspects of your application, it can be difficult to see the overall picture. Composite Evaluators...
Middleware in LangChain 1.0 alpha
Middleware in LangChain 1.0 gives you fine-grained control around the core agent loop of models + tools. It includes: Read the blog about middleware and try...
Create org-scoped API keys with granular permissions
You can now create org-scoped service keys and optionally, assign roles for created keys. These updates give you more control, allowing you to: Learn more...
Revision queueing in LangSmith Deployment
Keep your deployments smooth and predictable. With Revision Queueing , any new revisions are automatically lined up and processed only after the current one...
LangSmith Self-Hosted v0.11
This release brings customizable LangGraph Platform deployments, streamlined evaluation via Align Evals, running evals directly from Studio, and operational...
Trace Mode in LangSmith Studio
Debugging and improving agents just got a lot smoother. With Trace Mode , you can now view your LangSmith traces directly in Studio —no more jumping between...
Dynamic tool calling in LangGraph agents
Agents don’t always need the same tools at every step. With dynamic tool calling , you can now control which tools are available at different points in a...