Experiments in LangGraph Studio Show “Completed” but No Runs Available

LangSmith Product Help Observability & Evals
1

I am running experiments in LangGraph Studio through LangSmith. I have added datasets with examples and configured an LLM as the judge for correctness evaluation. The setup is for a RAG application.
When I execute the experiment against the staging environment via LangGraph Studio, the status shows “Experiments completed”, but no runs are displayed. That said, I can still see the traces under “Tracing Projects” as well as in “Evaluators”.

Experiment details

Screenshot 2025-09-30 at 14.14.12
Screenshot 2025-09-30 at 14.14.122362×468 41.8 KB

Packages installed

"@langchain/community": "^0.3.55",
"@langchain/core": "^0.3.68",
"@langchain/langgraph": "^0.4.9",
"@langchain/langgraph-checkpoint-postgres": "^0.1.1",
"@langchain/langgraph-sdk": "^0.0.107",
"@langchain/openai": "^0.6.7"
2

HI @SujayPrabhu thanks for reporting.

Do you see any errors in your deployment logs? The traces are showing up under the deployment’s tracing project, correct? Additionally, do you have any custom tracing setup in your project?

3

Hi @arjun

  1. I do not see any errors in deployment logs. All APIs are returning 200
  2. Yes, traces are showing up under the deployed project
  3. No, we do not have any custom tracing setup

I had restriction of uploading only one image for this question. I have the detailed explanation with screenshots here: Experiments in LangGraph Studio Show “Completed” but No Runs Available · langchain-ai/langgraph · Discussion #6225 · GitHub

4

Any updates on this @arjun ?

5

Hi @SujayPrabhu apologies for the delay here. I wanted to confirm – is this for a local server? If so, do you have LANGSMITH_API_KEY correctly defined in your projects .env file?