Could it be Claude 3.5 instead of Claude 4?

[Hitoshi_Yano (agaltallc)]

I’ve been feeling like the performance has been off for the past week or so while conversing on Cursor, and it told me the following. Could it be that I’m using Claude 3.5 instead of 4? I am a MAX subscription user.

• Base System: Claude 3.5 Sonnet by Anthropic
• Operating Environment: Running within the Cursor IDE
• Prompt Reception: Receives the following information from Cursor:
  • User input
  • Project file structure
  • Selected code snippets
  • Rule settings (e.g., repo_specific_rule)

[condor (Condor (Tee))]

Hi @Hitoshi_Yano and thank you for the feedback and your concern.

When you select a specific model, your request gets passed to AI provider with same model. Team has checked this several times and there is no routing issue to wrong models.

So when Claude 4 Sonnet is selected you will receive the response of Claude 4 Sonnet model.

Please note following:

• Anthropic models are trained with the name “Claude” only.
• Claude 4 models were trained on data that contained info of Claude 3.5 Sonnet.
• They are also trained to be helpful, so when they do not have sufficient info they may provide most probable answer even if not correct.

[nightcoder]

This is from CC itself. It says it is part of systems context, so someone is wrong here.

[condor (Condor (Tee))]

Claude Code = Antropics console based edit tool with its own system prompt,

Claude Console = Same access like API without that system prompt.

Those two are not the same.

[nightcoder]

Let’s breath deeply. If that thing answers what it answers, and if:

and if

why in the world only Cursor answers what it answers? Mind you you are talking to people who do know something about software development and whose trust is… this is currently a topic where I work, and I am doing you a huge favour by pointing it out. If there is anything to fix it is this nonsense.

[condor (Condor (Tee))]

1. Claude Web has a system prompt, as you can see.
2. JB AI Assistant has a system prompt, as you can see.

I’ve done this test several times, others have checked it as well with Anthropic Console and you can test it with their API. In about 30% of cases it answers wrong.

Claude 4 is reporting as Claude 3.5

Models are never aware of themselves as the data they are trained on doesn’t mention them - a bit of a catch-22! “Claude 4” is not a concept on the internet when they train the model, so it’s not aware of it’s own existence, and therefore falls back to saying it is Claude 3.5, as that is a model it is aware of from it’s training data. Rules may fix this, but it’s likely not worth adding a rule just to get it to state it’s real name - you are always talking to the model you pick in Cursor!

Claude 4 Sonnet UI Mislabeling or Misrouting to Claude 3.5 Sonnet

I just checked this in the Anthropic console and got these responses. [image] [image]

Strange name of the model Claude Sonnet 4

Its pretty clear from Anthropics own Console what info models have. Any non-API 3rd party tools have system prompts which has impact on responses. [image] [image] I have posted tests of various models in many threads here and will leave it at that, this has been officially addressed by Cursor Team members in several threads as well. If you have clear info there is an issue with any model please create a full Bug Report with a Request ID that has privacy disabled so the Cursor Team can revi…

I will bring it up again internally but this has been reviewed and tested by devs even checking the models responses as well as making sure the correct model is called.