3 ms·

But if you try some penny-saving cheap model like Sonnet [..bad things..]. [Better] pay through the nose for Opus. After blowing $800 of my bootstrap startup f

by dagss 8mo ago

But if you try some penny-saving cheap model like Sonnet [..bad things..]. [Better] pay through the nose for Opus. After blowing $800 of my bootstrap startup funds for Cursor with Opus for myself in a very productive January I figured I had to try to change things up... so this month I'm jumping between Claude Code and Cursor, sometimes writing the plans and having the conversation in Cursor and dump the implementation plan into Claude. Opus in Cursor is just so much more responsive and easy to talk to, compared to Opus in Claude. Cursor has this "Auto" mode which feels like it has very liberal limits (amortized cost I guess) that I'm also trying to use more, but -- I don't really like to flip a coin and if it lands up head then waste half hour discovering the LLM made a mess the LLM and try again forcing the model. Perhaps in March I'll bite the bullet and take this authors advice.

written-beyond 8mo ago
Just use Codex 5.3 in codex cli, the $20/mo plan is basically limitless at least for me and I keep reasoning efforts high. You can enjoy it while it lasts, OpenAI is being very liberal with their limits because of CC eating their lunch rn.
- rmonvfer 8mo ago
  Yeah, I can’t recommend gpt-5.3-codex enough, it’s great! I’ve been using it with the new macOS app and I’m impressed. I’ve always been a Claude Code guy and I find myself using codex more and more. Opus is still much nicer explaining issues and walking me through implementations but codex is faster (even with xhigh effort) and gets the job done 95% of the time. I was spending unholy amounts of money and tokens (subsidized cloud credits tho) forcing Opus for everything but I’m very happy with this new setup. I’ve also experimented with OpenCode and their Zen subscription to test Kimi K2.5 an similar models and they also seem like a very good alternative for some tasks. What I cannot stand tho is using sonnet directly (it’s fine as a subagent), I’ve found it to be hard to control and doesn’t follow detailed instructions.
  - jorl17 8mo ago
    Out of curiosity, what’s your flow? Do you have codex write plans to markdown files? Just chat? What languages or frameworks do you use? I’m an avid cursor user (with opus), and have been trying alternatives recently. Codex has been an immense letdown. I think I was too spoiled by cursor’s UX and internal planning prompt. It’s incredibly slow, produces terribly verbose and over-complicated code (unless I use high or xhigh, which are even slower), and missed a lot of details. Python/django and react frontend. For the first time I felt like I could relate to those people who say it doesn’t make them faster,” because they have to keep fixing the agent’s shot, never felt that with opus 4.5 and 4.6 and cursor
    - written-beyond 8mo ago
      Codex cli is a very performant cli though, better than any other cli code assistant I've used. I mean does it matter what code it's producing? If it renders and functions just use it. I think it's better to take the L on verbose code and optimizing the really ugly bits by hand in a few minutes than be kneecapped every 5 hour by limits and constant pleas to shift to Sonnet.
  - drsalt 8mo ago
    you've always been a Claude Code guy? this has existed less than a year.
    - rl3 8mo ago
      To be fair that still feels like an eternity somehow. Perhaps AI time is the inverse of Valve time.
    - redanddead 8mo ago
      I was born clutching a Claude Code shell, you peasant. The first sentence out of my mouth was a system prompt
- baq 8mo ago
  +1, codex 5.2 was really good and 5.3 seems to be better at everything; caveat - I had little time to test it.
- dagss 8mo ago
  Thanks a lot, after today I have fully switched to Codex I think. This vscode extension makes it almost as easy to point codex to something as when doing it in cursor: https://github.com/suzukenz/vscode-copy-selection-with-line-numbers/blob/main/.vscode/tasks.json https://github.com/suzukenz/vscode-copy-selection-with-line-...
dakolli 8mo ago
I promise you you're just going to continue to light money on fire. Don't fall for this token madness, the bigger your project gets, the less capable the llm will get and the more you spend per request on average. This is literally all marketing tricks by inference providers. Save your money and code it yourself, or use very inexpensive llm methods if you must. I think we are going to start hearing stories of people going into thousands in CC debt because they were essentially gambling with token usage thinking they would hit some startup jackpot.
- otabdeveloper4 8mo ago
  You can rent a GPU server and run your own Qwen models. It's 90 percent the same thing as Claude but with flat-rate costs.
- dagss 8mo ago
  Compared to the salary I loose by not taking a consulting gig for half a year, these $800 arent't all that much. (I guess depending on definition of bootstrap, mine might not be, as I support myself with saved consulting income.) Startup is a gamble with or without the LLM costs. I have been coding for 20 years, I have a good feel for how much time I would have spent without LLM assistance. And if LLMs vanish from the face of the earth tomorrow, I still saved myself that time.
  - dakolli 8mo ago
    Well if you've been "coding" for 20 years you should have known after a pretty short period of time it doesn't save you any time.

by dagss 8mo ago

written-beyond 8mo ago

Just use Codex 5.3 in codex cli, the $20/mo plan is basically limitless at least for me and I keep reasoning efforts high. You can enjoy it while it lasts, OpenAI is being very liberal with their limits because of CC eating their lunch rn.

rmonvfer 8mo ago
Yeah, I can’t recommend gpt-5.3-codex enough, it’s great! I’ve been using it with the new macOS app and I’m impressed. I’ve always been a Claude Code guy and I find myself using codex more and more. Opus is still much nicer explaining issues and walking me through implementations but codex is faster (even with xhigh effort) and gets the job done 95% of the time. I was spending unholy amounts of money and tokens (subsidized cloud credits tho) forcing Opus for everything but I’m very happy with this new setup. I’ve also experimented with OpenCode and their Zen subscription to test Kimi K2.5 an similar models and they also seem like a very good alternative for some tasks. What I cannot stand tho is using sonnet directly (it’s fine as a subagent), I’ve found it to be hard to control and doesn’t follow detailed instructions.
- jorl17 8mo ago
  Out of curiosity, what’s your flow? Do you have codex write plans to markdown files? Just chat? What languages or frameworks do you use? I’m an avid cursor user (with opus), and have been trying alternatives recently. Codex has been an immense letdown. I think I was too spoiled by cursor’s UX and internal planning prompt. It’s incredibly slow, produces terribly verbose and over-complicated code (unless I use high or xhigh, which are even slower), and missed a lot of details. Python/django and react frontend. For the first time I felt like I could relate to those people who say it doesn’t make them faster,” because they have to keep fixing the agent’s shot, never felt that with opus 4.5 and 4.6 and cursor
  - written-beyond 8mo ago
    Codex cli is a very performant cli though, better than any other cli code assistant I've used. I mean does it matter what code it's producing? If it renders and functions just use it. I think it's better to take the L on verbose code and optimizing the really ugly bits by hand in a few minutes than be kneecapped every 5 hour by limits and constant pleas to shift to Sonnet.
- drsalt 8mo ago
  you've always been a Claude Code guy? this has existed less than a year.
  - rl3 8mo ago
    To be fair that still feels like an eternity somehow. Perhaps AI time is the inverse of Valve time.
  - redanddead 8mo ago
    I was born clutching a Claude Code shell, you peasant. The first sentence out of my mouth was a system prompt
baq 8mo ago
+1, codex 5.2 was really good and 5.3 seems to be better at everything; caveat - I had little time to test it.
dagss 8mo ago
Thanks a lot, after today I have fully switched to Codex I think. This vscode extension makes it almost as easy to point codex to something as when doing it in cursor: https://github.com/suzukenz/vscode-copy-selection-with-line-numbers/blob/main/.vscode/tasks.json https://github.com/suzukenz/vscode-copy-selection-with-line-...

dakolli 8mo ago

I promise you you're just going to continue to light money on fire. Don't fall for this token madness, the bigger your project gets, the less capable the llm will get and the more you spend per request on average. This is literally all marketing tricks by inference providers. Save your money and code it yourself, or use very inexpensive llm methods if you must. I think we are going to start hearing stories of people going into thousands in CC debt because they were essentially gambling with token usage thinking they would hit some startup jackpot.

otabdeveloper4 8mo ago
You can rent a GPU server and run your own Qwen models. It's 90 percent the same thing as Claude but with flat-rate costs.
dagss 8mo ago
Compared to the salary I loose by not taking a consulting gig for half a year, these $800 arent't all that much. (I guess depending on definition of bootstrap, mine might not be, as I support myself with saved consulting income.) Startup is a gamble with or without the LLM costs. I have been coding for 20 years, I have a good feel for how much time I would have spent without LLM assistance. And if LLMs vanish from the face of the earth tomorrow, I still saved myself that time.
- dakolli 8mo ago
  Well if you've been "coding" for 20 years you should have known after a pretty short period of time it doesn't save you any time.