Could not parse response. Error when running Reflex app

[sinatijani]

Fails to parse.

Steps to reproduce

1. Cloned the repository: git clone https://github.com/reflex-dev/reflex-llm-examples.git
2. Navigated to the project directory.
3. Created and activated a virtual environment.
4. Installed the project's requirements.
5. Installed Playwright: playwright install. (Reflex wouldn't run so had to manually install this)
6. Ran the app: reflex run

Then in Browser use, i give the sample command given in the project:

1. Go to https://arxiv.org
2. Search for "Artificial Intelligence" or browse AI-related categories
3. Identify the latest papers (published in the last 7 days)
4. Summarize the title, authors, abstract, and publication date for each paper

After running, the following error shows up in the terminal:
(see the screenshot for the full terminal message)

INFO [agent]
📍 Step 1
ERROR [agent] ❌ Result failed 1/3 times:
Could not parse response.
INFO [agent]
📍 Step 1
INFO [agent] 🤷 Eval: Unknown - No previous actions to evaluate.
INFO [agent] 🧠 Memory:
INFO [agent] 🎯 Next goal: Start browser and navigate to arxiv.org
INFO [agent]
📍 Step 2
INFO [agent] 🤷 Eval: Unknown - No previous actions to evaluate.
INFO [agent] 🧠 Memory:
INFO [agent] 🎯 Next goal: Start browser and navigate to arxiv.org
INFO [agent]
📍 Step 3
ERROR [agent] ❌ Result failed 1/3 times:
Could not parse response.
INFO [agent]
📍 Step 3
ERROR [agent] ❌ Result failed 2/3 times:
Could not parse response.
INFO [agent]
📍 Step 3
ERROR [agent] ❌ Result failed 3/3 times:
Could not parse response.
ERROR [agent] ❌ Stopping due to 3 consecutive failures
INFO [watchfiles.main] 1 change detected
INFO [agent] Created GIF at agent_history.gif

I hope this helps to identify the issue to be fixed.

.

[Sumanth077]

Hey @sinatijani, The issue might be with the model, please try gemini or other models. Here is the example for you to checkout: https://docs.browser-use.com/customize/supported-models#gemini