story

Ask HN: Alternatives to Claude (Code)?

14 pointsvixalien1mo ago13 comments

Hello all, been trying to switch away from Claude Code and have been trailing this:

- Harness: Opencode (via Openchamber) - Subscription: GitHub Copilot (50$) - API Usage (beyond subscription): Open router - Free models: Opencode go

Here's the models I've trialed and like:

- Large (alternative to Opus): GPT 5.3 Codex - Medium (alternative to Sonnet): Minimax 2.7 - Smol: GPT 5.4 mini

these models are not yet on par to their respective Claude alternatives for me, but maybe I need to switch to something else or tune them, what have you been using?

Also looking for nice Opencode plugins, skills, completion providers (alternative to Cursor that I could use with Zen).

Happy to hear about your choices and experience

13 comments

SunshineTheCat1mo ago

I ping pong back and forth between claude code and codex.

In my experience (very subjective, obviously) for backend/"logical" tasks Codex seems to outperform Claude.

For front-end/UX related tasks Claude wins easily.

Overall, Claude does seem to be a little better in other areas too.

Codex's biggest advantage in my personal opinion, however, is usage. I think maybe once in several months did I even get close to hitting my limit with the $20 plan.

With Claude, however, I feel like I can sneeze and half my weekly usage is gone. Same $20 price tag.

That's been my experience, I'm sure it differs user to user though.

PocketBot1mo ago

I have never heard someone complimenting claude on frontend lol. To each their own though, fair play.

vixalienOP1mo ago

nice, I have been hearing about Codex a lot, so I will def try it

verdverm1mo ago

I've gone fully custom and would never look back. All of the popular tools lack the experience I'm after, generally lacking the advanced customization I have found makes weaker models more capable. I'm also moving towards open weight models.

The underlying motivation is to avoid Big Ai as much as possible. Incentives are not aligned.

kingkongjaffa1mo ago

> the advanced customization I have found makes weaker models more capable.

Like what?

verdverm1mo ago

Good prompts and tools. When I first went down this path, I grabbed the Claude Code prompts and adapted them. Doing so made gemini-3-flash as good as pro for 80%+ of my normal tasks. I became spoiled with the turnaround time (much faster) and now become impatient with the mega models.

I can also do things like out the AGENTS.md contents in the system prompt and change file read to put contents there as well. Thai avoids duplicates in the messages when they fail at editing and decide to reread the file

square_usual1mo ago

Just pay the $20 for codex and use it. It's the only real alternative.

omertt271mo ago

I am using both claude code and github copilot at same time.I like using copilot actually

zplb1mo ago

Perhaps you could try GLM5.1, I see the offical demo, it's nice

sminchev1mo ago

GLM, yes I use it, mostly for code review. I do code review, take the analysis and pass it to Opus for implementation, Relatively cheap, has good statistics

hybirdss1mo ago

GLM works well, but since it’s a Chinese model, you need to configure permissions carefully. This also applies when using it in the harness.

Dante777111mo ago

GLM plan with VS & RooCode

segmondy1mo ago

opencode, kimi-cli, qwen-cli, pi-agent,

j / k navigate · click thread line to collapse

13 comments

SunshineTheCat1mo ago

I ping pong back and forth between claude code and codex.

In my experience (very subjective, obviously) for backend/"logical" tasks Codex seems to outperform Claude.

For front-end/UX related tasks Claude wins easily.

Overall, Claude does seem to be a little better in other areas too.

Codex's biggest advantage in my personal opinion, however, is usage. I think maybe once in several months did I even get close to hitting my limit with the $20 plan.

With Claude, however, I feel like I can sneeze and half my weekly usage is gone. Same $20 price tag.

That's been my experience, I'm sure it differs user to user though.

PocketBot1mo ago

I have never heard someone complimenting claude on frontend lol. To each their own though, fair play.

vixalienOP1mo ago

nice, I have been hearing about Codex a lot, so I will def try it

verdverm1mo ago

The underlying motivation is to avoid Big Ai as much as possible. Incentives are not aligned.

kingkongjaffa1mo ago

> the advanced customization I have found makes weaker models more capable.

Like what?

verdverm1mo ago

square_usual1mo ago

Just pay the $20 for codex and use it. It's the only real alternative.

omertt271mo ago

I am using both claude code and github copilot at same time.I like using copilot actually

zplb1mo ago

Perhaps you could try GLM5.1, I see the offical demo, it's nice

sminchev1mo ago

GLM, yes I use it, mostly for code review. I do code review, take the analysis and pass it to Opus for implementation, Relatively cheap, has good statistics

hybirdss1mo ago

GLM works well, but since it’s a Chinese model, you need to configure permissions carefully. This also applies when using it in the harness.

Dante777111mo ago

GLM plan with VS & RooCode

segmondy1mo ago

opencode, kimi-cli, qwen-cli, pi-agent,

j / k navigate · click thread line to collapse