Hey all,
Anthropic has just released Claude Opus 4.1, and the model is already available for you in Cursor to try! While its behaviour should be unchanged from 4.0, Anthropic reports a higher intelligence against benchmarks.
Note: Claude Opus is a very expensive model to run, and will likely consume a significantly higher portion of your monthly usage with each request. Please use it wisely to ensure your included usage is not significantly consumed.
You can learn more about the model here:
Fast as always
When gpt oss?
And yet another time, open-source models get ignored. Who would have thought..
Not usable with the new pricing.
Thank you for your feedback, we may consider it if there is enough community boycott.
OMG I wish I could use it in non max mode
Not for me - non of the models except Auto are working for me for a third day and I have no support. Mailing [email protected] is pointless. “If the problem persists, please check your internet connection or VPN, or email us at [email protected].
(Request ID: 683729b4-5fe0-40e6-9fc9-a8ff14f26c33)”
I’m struggling also 2 days with all models in any scenario, directly in browser, in cursor. The all are giving worst codes. No matter what my prompt is, the answers are always ‘this will fix it’, hahaha. Last week was much better. No clue, LLM are volatile ever since it came to this world.
I would like to know exactly how much its going to drain my Ultra account to even try Opus 4.1 for 1 prompt.
Meanwhile spending the day here salvaging damaged code from yesterday’s bloodbath … killed the remaining chat sessions i carried over from yesterday just now. I have started using sonnet 3.7 again after so many foolish agentic acts of treason with sonnet 4
o3 removed permanently from my LLM list for trying to incorporate “middleware” python libraries - out right spyware implant attempts.
as a fellow FinDev’er, I’ll see you on the other end!
I prefer 1.4. It’s better! 
If you use AI-interaction tool, you will have more than 1 prompt lol
never