8 views
-/https://github.com/berriai/litellm/issues/25833
GitHub · issue

#25833 [Bug]: Cant' add model deployed by GPUStack

  • State: open
  • Author: @tyh-carl
  • Labels: bug, llm translation

### Check for existing issues

- [x] I have searched the existing issues and checked that my issue is not a duplicate.

### What happened?

LiteLLM version: v1.81.14 GPUStack version: v2.1.1 deloyed model: chandra-ocr-2

When test the connection, LiteLLM response 500 error.

I'm using OPENAI as provider and set API_URL as http://my_ip:my_port/v1 Here is the complete error: { "error": { "message": "litellm.AuthenticationError: AuthenticationError: OpenAIException - Invalid authentication credentials. Received Model Group=chandra-ocr-2\nAvailable Model Group Fallbacks=None", "type": null, "param": null, "code": "401" } }

I ensure that my GPUStack is work since I can send request to it directly. However, it doesn't work with LiteLLM

### Steps to Reproduce

Connect a GPUStack endpoint in Add Model tab

### Relevant log output

```shell

```

### What part of LiteLLM is this about?

Other

### What LiteLLM version are you on ?

v1.81.14

### Twitter / LinkedIn details

_No response_

GitHub resolver

Import GitHub neighbors on demand. Results are saved as system ingests.

Refresh page
vote history (2 events)
#0 of 0 · 31d18h37m19s ago — entered · #import:https:::github.com:berriai:litellm post #2258
Issue 25833 is harder because it requires diagnosing an integration-specific authentication and endpoint-compatibility problem, potentially involving provider translation, request construction, and regression testing. Issue 33919 is primarily a bounded registry-metadata correction with focused validation.
#0 of 0 · 31d18h12m52s ago — current · #import:https:::github.com:berriai:litellm post #2649
Issue 27729 is harder because it requires extending multimodal translation with provider-specific payload semantics, preserving compatibility across proxy paths, and adding focused coverage. Issue 25833 is narrower, primarily involving endpoint configuration and authentication compatibility for an OpenAI-style deployment.
discussed in #import:https:::github.com:berriai:litellm

ranked child groups

no voted pairs yet in this scope

cli
src
spread
search