[Bug] Unable to generate chat title when `CUSTOM_MODELS` removed `gpt-3.5-turbo`

[Cow258]

Describe the bug

When CUSTOM_MODELS removed gpt-3.5-turbo, Chat title will never be generated.

To Reproduce

Steps to reproduce the behavior:

1. Set env CUSTOM_MODELS=-gpt-3.5-turbo
2. Deploy
3. Create new chat
4. Ask something
5. GPT will response successfully
6. but chat title will never be generated

Expected behavior

Even if gpt-3.5-turbo disabled, the chat title should always be generated.

Screenshots

POST - /api/openai/v1/chat/completions HTTP Status: 403

{
  "error": true,
  "message": "you are not allowed to use gpt-3.5-turbo model"
}

Deployment

• [ ] Docker
• [x] Vercel
• [ ] Server

Desktop (please complete the following information):

• OS: Any
• Browser: Any
• Version: Any

Smartphone (please complete the following information):

• Device: Any
• OS: Any
• Browser: Any
• Version: Any

Additional Logs Add any logs about the problem here.

[H0llyW00dzZ]

been known this issue since a few month ago it's because of this code

function getSummarizeModel(currentModel: string) {
  // if it is using gpt-* models, force to use 3.5 to summarize
  return currentModel.startsWith("gpt") ? SUMMARIZE_MODEL : currentModel;
}

location: chat.ts

on my fork, working fine now without that bug

[H0llyW00dzZ]

this better now I've been pull req before (a few month ago but yiidaa won't merge)

• Fix [UI/UX Page] Known issue #3624