Hey everyone,
I’ve been testing the new versions of Gemini (both MAX and the one included in the $20/month plan) for a while now, and honestly… I’m not noticing any real improvements. In fact, sometimes it performs worse than the previous version.
Even though many say Gemini is solid, I’m starting to think it might not be working at full potential on Cursor. Hopefully, with the upcoming v0.50.0 update, we’ll see improvements—especially in these two key areas:
- Function calling: it often fails to properly edit files or analyze the input resources I provide. Sometimes it doesn’t even search or extract key elements correctly.
- Latency: editing a file with 500 lines of code can take 2–3 minutes. That shouldn’t happen, especially with the performance boost we’ve seen in recent benchmarks.
I really hope this gets addressed soon. If you’re running into similar issues, let me know — it’d be great to discuss and maybe share ideas to improve this.
Also, I believe it’s crucial that the Cursor team prioritizes Gemini support over Claude, since Gemini seems to be gaining more traction in the market.
Thanks!