What is the context size for gpt 4o mini?

[Niko_Bellic]

What is the context size for GPT 4o mini?
I have to use it a lot because I am trying to save 500 requests for the most important task…
So is it 20k tokens?

[litecode]

I would be interested to know:

01) The actual context size of the LLMs available, ie:

• Normal Chat:

  • cursor-small
  • gpt-3.5
  • gpt-4o-mini
  • gpt-4o
  • claude-3.5-sonnet

• Long Context Chat:

  • gpt-4o-128k
  • gemini-1.5-flash-500k
  • claude-3-haiku-200k
  • claude-3.5-sonnet-200k

02) The context size of the LLMs when used within Cursor

There seems to be a bit of uncertainty about this:

Gpt4-0, context size?

Note the model is capable of 128k context, but cursor only uses 10k Why are the models on using your own keys so much better? - #6 by arvid220u, not sure how much of that is input/output.

Why are the models on using your own keys so much better?

No, we’re using 10k tokens regardless of model.

Request: Allow to set own context size limit

Hello! I want to ask the team to provide a way to disable the 10k limit for context window, possibly allowing us to set our own limits.

GPT-4o in Long Context Mode

The 4 Turbo and 4o context length is 20k in chat for now (might change with 4o’s speed)!

And I saw you recently asked a question about the general topic as well:

Please allow max number of tokens supported by the models

Can devs be more transparent with token limits for PRO for different models? In many cases, it could be a time saver - at least you could know that the size of the current context is too big to discuss for the current LLM and save your time (and requests) not doing that. As long as it is not obvious and devs don’t talk about limits much (I can’t find it on site, on docs, on FAQ, etc) you start to think it’s pretty small? I’ve heard something about 8k/10k but it could be changed, there are new mo…

I am a bit confused why there would be long-context models available to select in Cursor, if they weren’t actually long-context models when utilized in Cursor.

Please excuse any ignorance if I am not understanding the idea of a ‘long-context’ model correctly.

[Niko_Bellic]

From Cursor - Built Software Faster
It seems like it should be 20k but I am really not sure.

What context window is used for model X?

In chat, we limit to around 20,000 tokens at the moment (or less if the model does not support that much context). For cmd-K, we limit to around 10,000 tokens, to balance TTFT and quality. Long-context chat uses the model’s maximum context window.