Add Support for gpt-4-turbo-2024-04-09

[SlickyNicky]

New model release: https://platform.openai.com/docs/models/gpt-4-turbo-and-gpt-4

[Clemens123]

Hi, I'm not exactly sure why the pull request mentioned above was closed, but will gpt-4-turbo-2024-04-09 become available in chatgpt-web? It's supposed to be a substantial improvement and it would be a pitty if there were technical reasons preventing that :(

[PrimalOutshoot]

I have tried adding 'gpt-4-turbo': { ...gpt4128kpreview }, to export const chatModels : Record<string, ModelDetail> in the src > lib > providers > openai> models.svelte

It has a context window of 128k same as the preview, but for some reason I get an Error 200 - Unknown Response. The problem is when I add gpt-4-turbo to the chatModels function, using these settings:

const gpt4128kpreview = { ...chatModelBase, prompt: 0.00001, // $0.01 per 1000 tokens prompt completion: 0.00003, // $0.03 per 1000 tokens completion max: 131072 // 128k max token buffer }

I get an Error 200 with the gpt4-turbo model

export const chatModels : Record<string, ModelDetail> = { 'gpt-4-turbo': { ...gpt4128kpreview }, 'gpt-4-turbo-preview': { ...gpt4128kpreview } }

gpt-4-turbo should in theory work with the same settings as gpt-4-turbo-preview, but it throws back an "Error 200" instead, which I have not been able to resolve on my end due to lack of knowledge in the area. I have been waiting for someone more knowledgeable than me to throw a solution in chatgpt-web, but it has not been as simple for me to add gpt-4-turbo as I have added other models manually in the past. I am not sure what setting it is supposed to have, or why it is not working, and search has not been fruitful. Would love to see support or working implementation added for gpt-4-turbo.

[tikkamasalla]

same, been waiting for it for a while

[Niek]

Fixed in #448