[GH-ISSUE #2925] [BUG]: Generic OpenAI endpoint does not work with Ollama #1866

Closed
opened 2026-02-22 18:26:53 -05:00 by yindo · 1 comment
Owner

Originally created by @sharpe5 on GitHub (Jan 1, 2025).
Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/2925

How are you running AnythingLLM?

AnythingLLM desktop app on MacOS (Macbook Pro).

What happened?

I want to connect AnythingLLM to Ollama using the generic OpenAI-compatible endpoint. I configure as per the API docs on Ollama:

image

However, when I create a new workspace, chats are broken.

Yes, it is possible to configure AnythingLLM to talk directly to Ollama, but this is also broken and I would prefer to use a generic OpenAI-compatible endpoint.

Are there known steps to reproduce?

To reproduce:

  1. Install ollama locally from www.ollama.com.
  2. Run Ollama locally:
gravitas@PHIENGINE1 C:\Users\Gravitas>ollama run llama3.2
>>> test
It looks like you're just testing to see if I'm working.
  1. Attempt to configure AnythingLLM to talk to Ollama via the OpenAI-compatible endpoint.
Originally created by @sharpe5 on GitHub (Jan 1, 2025). Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/2925 ### How are you running AnythingLLM? AnythingLLM desktop app on MacOS (Macbook Pro). ### What happened? I want to connect AnythingLLM to Ollama using the generic OpenAI-compatible endpoint. I [configure as per the API docs on Ollama](https://github.com/ollama/ollama/blob/main/docs/openai.md): <img width="831" alt="image" src="https://github.com/user-attachments/assets/5b72949c-0b0c-45a5-8bfa-3a79660178f0" /> However, when I create a new workspace, chats are broken. Yes, it is possible to configure AnythingLLM to talk directly to Ollama, but this is also broken and I would prefer to use a generic OpenAI-compatible endpoint. ### Are there known steps to reproduce? To reproduce: 1. Install ollama locally from [www.ollama.com](www.ollama.com). 2. Run Ollama locally: ``` gravitas@PHIENGINE1 C:\Users\Gravitas>ollama run llama3.2 >>> test It looks like you're just testing to see if I'm working. ``` 3. Attempt to configure AnythingLLM to talk to Ollama via the OpenAI-compatible endpoint.
yindo added the possible bug label 2026-02-22 18:26:53 -05:00
yindo closed this issue 2026-02-22 18:26:53 -05:00
Author
Owner

@timothycarambat commented on GitHub (Jan 1, 2025):

Yeah, because it works differently than ollama does for stream chunk handling. Just use the Ollama provider since it is fully supported as an LLM anyways - then you wont have any issues!

@timothycarambat commented on GitHub (Jan 1, 2025): Yeah, because it works differently than ollama does for stream chunk handling. Just use the Ollama provider since it is fully supported as an LLM anyways - then you wont have any issues!
yindo changed title from [BUG]: Generic OpenAI endpoint does not work with Ollama to [GH-ISSUE #2925] [BUG]: Generic OpenAI endpoint does not work with Ollama 2026-06-05 14:43:08 -04:00
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Mintplex-Labs/anything-llm#1866