· Add/Enable o3 Pro support |OpenAI Provider| in WindSurf Cascade 12.5 · =)
ARTISTIC AUDIO
OpenAI’s o3 Pro and ChatGPT-5 Pro are much better at math. Claude 4.1 costs 20 credits and performs worse than Gemini 2.5 Pro or GPT-5 Thinking.
o3 Pro — key use cases: PhD-level math function optimization; scientific analysis; math-heavy code synthesis/refactoring/optimization.
There is strong demand across Windsurf and AI-dev Discord, Facebook, and Slack rooms for o3 Pro in Cascade for math/science/code workflows (e.g., PhD-level math function optimization, scientific analysis, math-heavy code synthesis).
Once available, please also include the ChatGPT-5 Pro model.
I’m happy to share reproducible prompt sets and benchmarks, and to help as an early tester to validate output quality.
Thanks!
Clarissa Bella
Thanks this shared suggestion/feedback link in my private scientific discord channel.
In most cases I use only o3 Pro and GPT 5 Pro Thinking in my lab (R&D). o3 Pro, Gemini 2.5 Pro and GPT 5 Thinking can solve research-level math. Claude 4.1 is useless in this level, but I like 4.1 for lightweight tasks, SEO projects, etc.
lswiss dev
Cool idea, cheers from Discord lab :P
Grayson “Grayson M” M
Yes, OpenAI’s o3 Pro, Gemini 2.5 Pro, and GPT 5 Thinking/Pro also beats Claude Opus 4.1 in math/science. Significant difference.
Iguana CGI
+1 I love o3 Pro in math. e.g.: OpenRouter in Kilo Code.