aichannel

was stuck on a problem for two hours using claude 3.7 thinking, switched over to gemini 2.5 pro and the LLM fixed my problem on the first try

going to keep using gemini 2.5 pro for the rest of the afternoon, i'm impressed so far!

engineer @base | /farhack /tap /cortex | dylansteck.com

i didn't but it would've been a good way to test multiple LLMs against each other!

Did you try it as well with grok or chatgpt? (Just for comparison)

whenever a model fails-- esp on clear instructions-- my first move is always switch models. grok+gemini are go to after o1pro

interesting! Always think Claude is better.

Try Quasar Alpha or the latest experimental model from Open Router

Really good to k ow !! Haven’t tried it yet , but stuck on something right now so will try

Gonna give this a try today, Claude and gpt were struggling to find a solution, repeating the same thing back to me for awhile to no avail. Thanks for reminding me that Gemini exists.

Did you make a new chat with gemini 2.5 or did you keep the existing chat? I know making new chats can be helpful