I used to use Gemini CLI and was pretty happy with it… until I started using Claude Code. I think it’s, for now, the most advanced AI tool available right now. You can manage skills, the MCP, plugins, agents, hooks… It’s a truly comprehensive tool, but its main drawback is the token consumption of the Anthropic model, especially in recent days. There are solutions to reduce input tokens, for example this GitHub project https://github.com/rtk-ai/rtk. It uses hooks to replace certain commands with optimized ones and helps reduce the number of input tokens. It’s a start...

Thanks, anyways I have the Google AI Pro subscription for one year, so for now I will continue using Gemini

I love Gemini for general knowledge and think it far outperforms Claude for research, but personally I cannot use Gemini for programming due to: 1. it taking so long 2. it not giving any information about its thought process / what it's doing "in its head", whereas Claude code tells you what it's thinking and allows you to interrupt.

So for now, I would personally recommend dropping Gemini for the actual implementation of programming or anything involving analysis of your codebase.

I don't doubt that Google has smart people on this, and that this may improve in a few months' time.

I understand your position, however for some strange reason (maybe because like I am the kind of software engineer who liks to do most by himself) I haven't noticed any issue every time I ask Gemini to do something. Thanks anyways for your comment!!!

I haven't used Gemini CLI, but when I'm using Claude I use Opus for planning, and Sonnet for executing things. I barely use Haiku, but I sometimes use it for brainstorming or really simple tasks.

I think Claude is far better than Gemini, the lack of usage is a chronic problem. Even using the Max model is not enough. If I used that much, I think I worked a lot and took a rest. LOL.

claude make the plan, and let gemini implement.

Interesting POV