[Done] Request GPT-4o model on cursor

As an avid user of cursor.ai and someone who closely follows developments in AI and natural language processing, I believe adding GPT-4o would significantly enhance the capabilities and user experience of cursor’s service.

GPT-4o represents a major leap forward in language model performance. Its advanced architecture and training on a vast corpus of data enable it to engage in open-ended conversation, answer follow-up questions, and assist with an impressive range of writing and analytical tasks. Cursor.ai has always been at the forefront of providing cutting-edge AI tools, and I think GPT-4o would be a natural and exciting addition to your offerings.

Some key benefits of GPT-4o that I believe would bring great value to cursor.ai users include:

  1. Improved language understanding and generation across a wide range of domains
  2. Enhanced ability to maintain coherence and context over long conversations and complex, multi-step tasks
  3. More nuanced and insightful responses that draw upon the vast knowledge the model has acquired
  4. Stronger performance on analytical and reasoning tasks like strategic planning, code explanation, and mathematical problem-solving

I know that incorporating a powerful new model like GPT-4o is no small undertaking, but I’m confident it would pay dividends in terms of increased user engagement, expanded use cases, and positioning cursor.ai as the go-to platform for the most advanced publicly available language model.

Please consider this request to introduce GPT-4o, and let me and the rest of the cursor.ai community know if there are any plans to add this impressive model to the platform. I’m excited to see cursor.ai continue to evolve and push the boundaries of what’s possible with language AI.

2 Likes

Already on my cursor editor. I wonder what the limits will be. Since it’s much cheaper, i hope atleast 1000 gpt-4o calls for cursor pro users

6 Likes

+1 That would be awesome! :smiley:

1 Like

GPT-4o is up! If you reload your editor, you should see it in the model dropdown. If not, you can punch it in manually in Settings > Models.

Still figuring out pricing. For now, gpt-4o requests will just count as gpt-4 requests.

6 Likes

It would be most interesting to use gpt4o in the long context mode!

1 Like

Yea, is it better at that like opus?

opus are much slower and very expensive! Encouraging the see the high score gpt4o banged on coding so it would be most interesting to try it.
via api I have seen gpt4o to be enable to keep long coherence up to 50k context length.

1 Like

does the new GPT-4o model have 2024 knowledge? or is it still cut off in 2022?

  • Turbo is up to Dec 2023 not 2022
  • GPT-4o is up to Oct 2023

https://platform.openai.com/docs/models/gpt-4o

2 Likes

It appears to be worse at hard tasks:
Screenshot 2024-05-14 at 03.47.29
image

1 Like

Sad but almost expected somehow… Given its smaller size, there are probably trade-offs in performance when handling challenging, “unseen” cases where the model needs to generalize.

I think the most interesting aspects of this model are not its performance but its price, speed, and real-time multimodal capabilities.

2 Likes

Thanks for getting this integrated so quickly. Hugely appreciated :fire:

1 Like

Is it possible to enable the interpreter mode for gpt-4o? Right now only gpt4 and gpt-4-turbo works

Why can’t API key user get the UI option to use gpt-4o???

I have updated my cursor and have provided my own OpenAI API key, but the gpt-4o model isn’t showing up. Any ideas?

1 Like

Does it work if you manually add gpt-4o?

Same thing happed to me.

I tried and asked in the chat to find out if the model was changed by asking the knowledge cutoff. It was 2021


It seemed to have changed.