←back to thread

Gemini CLI

(blog.google)
1342 points sync | 6 comments | 25 Jun 25 13:10 UTC | HN request time: 2.133s | source | bottom
Show context
asadm ◴[25 Jun 25 13:35 UTC] No.44377180[source]▶
I have been using this for about a month and it’s a beast, mostly thanks to 2.5pro being SOTA and also how it leverages that huge 1M context window. Other tools either preemptively compress context or try to read files partially.

I have thrown very large codebases at this and it has been able to navigate and learn them effortlessly.

replies(2): >>44377204 #>>44378419 #
1. _zoltan_ ◴[25 Jun 25 15:27 UTC] No.44378419[source]▶
what's your workflow?
replies(2): >>44382980 #>>44383035 #
2. leoh ◴[26 Jun 25 00:03 UTC] No.44382980[source]▶
+1 I have not found Gemini 2.5 better than Claude's latest models -- different and better at some things, but not better in general; and in general I have found Gemini 2.5 Pro to be worse at dealing with large codebases despite its huge context window. So I too am quite curious about the workflow here.
replies(1): >>44383039 #
3. asadm ◴[26 Jun 25 00:15 UTC] No.44383035[source]▶
i guess one core idea is to point it to an entrypoint or some part of code you are planning to touch. instead of asking it to load up every file etc. and then figure out the control flow.
4. asadm ◴[26 Jun 25 00:15 UTC] No.44383039[source]▶
the core idea is to not use up all of context for files but instead sessions go longer before becoming a useless pursuit ie. more turns possible with larger context window
replies(1): >>44385361 #
5. leoh ◴[26 Jun 25 08:27 UTC] No.44385361{3}[source]▶
Are you doing this with an LLM interface directly as opposed to using cursor or another tool like that?
replies(1): >>44389860 #
6. asadm ◴[26 Jun 25 18:15 UTC] No.44389860{4}[source]▶
yeah i meant via coding interfaces, mostly aider/claude code/gemini cli