Open-WebUI: ‘stream=true’ Mode Not Working with Ollama on RunPod #3049

Closed
opened 2025-11-11 15:20:33 -06:00 by GiteaMirror · 0 comments
Owner

Originally created by @mballesterosc on GitHub (Dec 19, 2024).

Installation Method

pip

Environment

  • Open WebUI Version: 0.4.8

  • **Ollama [v0.5.4] on RunPod

  • Operating System: Windows Server 2016

  • **Browser Chrome

Bug Summary:

Thought about stream mode issue durante un segundo
I have Open-WebUI installed on a local server with an Ollama installation on a RunPod server. The connection and inference work perfectly, but I can’t get the stream=True mode to function. I’ve tried connecting to Ollama both via HTTP and TCP (IP:PORT), and I’ve configured the HOST_OLLAMA environment variable to 0.0.0.0 on RunPod. The call correctly sends the stream=true parameter.

If I call the Ollama API from Python, it works correctly.
If I use “Arena Model” with only my RunPod Ollama model (LLAMA 3.3) available in Open-WebUI, streaming works correctly.
I’ve checked all the threads, asked RunPod, but haven’t gotten anywhere.

Could you help me or guide me in solving this issue? I’m not sure if it’s a bug in Open-WebUI, a RunPod issue, or something related to my local server.

Thank you in advance.

Originally created by @mballesterosc on GitHub (Dec 19, 2024). ## Installation Method pip ## Environment - **Open WebUI Version:** 0.4.8 - **Ollama [v0.5.4] on RunPod - **Operating System:** Windows Server 2016 - **Browser Chrome **Bug Summary:** Thought about stream mode issue durante un segundo I have Open-WebUI installed on a local server with an Ollama installation on a RunPod server. The connection and inference work perfectly, but I can’t get the stream=True mode to function. I’ve tried connecting to Ollama both via HTTP and TCP (IP:PORT), and I’ve configured the HOST_OLLAMA environment variable to 0.0.0.0 on RunPod. The call correctly sends the stream=true parameter. If I call the Ollama API from Python, it works correctly. If I use “Arena Model” with only my RunPod Ollama model (LLAMA 3.3) available in Open-WebUI, streaming works correctly. I’ve checked all the threads, asked RunPod, but haven’t gotten anywhere. Could you help me or guide me in solving this issue? I’m not sure if it’s a bug in Open-WebUI, a RunPod issue, or something related to my local server. Thank you in advance.
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/open-webui#3049