[Azure OpenAI] - Title generation fails + librechat crashes every message

[m-bertolino-exc]

What happened?

This problem happens on a clean install of librechat, deployed using docker compose on macos. I was able to reproduce it also on a k8s cluster(GKE).

My configuration

I did not modify anythning relevant in the .env.example file, I just renamed it as .env and added the azure openai endpoint.
My librechat.yaml is:(sensitive data redacted)

version: 1.2.1
endpoints:
  azureOpenAI:
    titleModel: "gpt-4o"
    titleConvo: true
    plugins: true
    groups:
    - group: "ic_models2"
      apiKey: "xxx"
      instanceName: "xxx"
      version: "xxx"
      models:
        gpt4o:
          deploymentName: "lund-gpt-4o"
          version: "2023-03-15-preview"

The problem

I create a new account, I select azure open ai, I select my model, I type "a" and press enter.
I receive a proper answer for the model, but on the networks tab I already see that 2 requests failed:

• http://localhost:3080/c/6455ae42-d1d1-4cc3-ab1e-2395ec2ac506
• http://localhost:3080/api/convos/gen_title

The status for both is (failed)net::ERR_EMPTY_RESPONSE

If I now open my terminal, I see that the Librechat container just crashed. The error is:
LibreChat | 2025-01-28 08:48:08 error: There was an uncaught error: Model named "gpt-4o" not found in configuration.

And the server restarts.
The chat also fails to get a generated title, as the server crashes before generating one.

Imporant note

This problem is not related with actually not being able to finding the model in the configuration, as I successfully get a response from the model. I also tried to play around with model names, it did not solve the issue.

Steps to Reproduce

1. clone the repository
2. add azure open ai to endpoints in .env
3. add a azure openai endpoint in librechat.yaml
4. docker compose up
5. register an account and send a message

What browsers are you seeing the problem on?

Chrome, Firefox

Relevant log output

2025-01-28T08:48:11.259Z info: [Optional] Redis not initialized. Note: Redis support is experimental.
2025-01-28T08:48:12.060Z info: Connected to MongoDB
2025-01-28T08:48:12.182Z debug: [indexSync] There are 32 messages and 32 indexed
2025-01-28T08:48:12.182Z debug: [indexSync] There are 13 convos and 13 indexed
2025-01-28T08:48:12.199Z info: Custom config file loaded:
2025-01-28T08:48:12.200Z info: {
  "version": "1.2.1",
  "endpoints": {
    "azureOpenAI": {
      "titleModel": "gpt-4o",
      "titleConvo": true,
      "plugins": true,
      "gr... [truncated]
2025-01-28T08:48:12.200Z debug: Custom config:
{
  version: "1.2.1",
    azureOpenAI.titleModel: "gpt-4o",
    azureOpenAI.titleConvo: true,
    azureOpenAI.plugins: true,
    // 1 group(s)
    azureOpenAI.groups: [{"group":"ic_models2","apiKey":"... [truncated]],
}
2025-01-28T08:48:12.201Z warn: Default value for CREDS_KEY is being used.
2025-01-28T08:48:12.202Z warn: Default value for CREDS_IV is being used.
2025-01-28T08:48:12.203Z warn: Default value for JWT_SECRET is being used.
2025-01-28T08:48:12.203Z warn: Default value for JWT_REFRESH_SECRET is being used.
2025-01-28T08:48:12.203Z info: Please replace any default secret values.
2025-01-28T08:48:12.204Z info: ​

    For your convenience, use this tool to generate your own secret values:
    https://www.librechat.ai/toolkit/creds_generator

    ​
2025-01-28T08:48:12.223Z info: RAG API is running and reachable at http://rag_api:8000.
2025-01-28T08:48:12.230Z info: No changes needed for 'USER' role permissions
2025-01-28T08:48:12.232Z info: No changes needed for 'ADMIN' role permissions
2025-01-28T08:48:12.234Z info: Custom config file loaded:
2025-01-28T08:48:12.235Z info: {
  "version": "1.2.1",
  "endpoints": {
    "azureOpenAI": {
      "titleModel": "gpt-4o",
      "titleConvo": true,
      "plugins": true,
      "gr... [truncated]
2025-01-28T08:48:12.235Z debug: Custom config:
{
  version: "1.2.1",
    azureOpenAI.titleModel: "gpt-4o",
    azureOpenAI.titleConvo: true,
    azureOpenAI.plugins: true,
    // 1 group(s)
    azureOpenAI.groups: [{"group":"ic_models2","apiKey":"... [truncated]],
}
2025-01-28T08:48:12.238Z info: Server listening on all interfaces at port 3080. Use http://localhost:3080 to access it
















{"level":"error","message":"There was an uncaught error: Model named \"gpt-4o\" not found in configuration.","stack":"Error: Model named \"gpt-4o\" not found in configuration.\n    at exports.mapModelToAzureConfig (/app/packages/data-provider/dist/index.js:1:194856)\n    at OpenAIClient.titleConvo (/app/api/app/clients/OpenAIClient.js:748:11)\n    at addTitle (/app/api/server/services/Endpoints/openAI/title.js:24:30)\n    at AskController (/app/api/server/controllers/AskController.js:170:7)\n    at process.processTicksAndRejections (node:internal/process/task_queues:95:5)\n    at async /app/api/server/routes/ask/openAI.js:24:5"}

Screenshots

Code of Conduct

• I agree to follow this project's Code of Conduct

[danny-avila]

The naming here is important:

      models:
        gpt4o:

It should match its corresponding official name i.e. gpt-4o