Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

A guide to local coding models

aiforswes.com

Post reply on HN
Print view View original

101–110 of 363 posts

Re: A guide to local coding models

#101

Post by bottlepalm »

simonw wrote: post #4

> I realized I looked at this more from the angle of a hobbiest paying for these coding tools. Someone doing little side projects—not someone in a production setting. I did this because I see a lot of people signing up for $100/mo or $200/mo coding subscriptions for personal projects when they likely don’t need to. Are people really doing that? If that's you, know that you can get a LONG way on the $20/month plans fr…

When you pay $1000/month for health insurance and $2000/month for housing.. $200 for something you actually enjoy isn't so bad.

Re: A guide to local coding models

#102

Post by satvikpendem »

maddmann wrote: post #44
satvikpendem wrote: post #27

Earlier quoted context omitted.

> If that's you, know that you can get a LONG way on the $20/month plans from OpenAI and Anthropic. > The time to cough up $100 or $200/month is when you've exhausted your $20/month quota and you are frustrated at getting cut off. At that point you should be able to make a responsible decision by yourself. These are the same people, by and large. What I have seen is users who purely vibe code everything and run into…

If this is the new way code is written then they are arguably learning how to code. Jury is still out though, but I think you are being a bit dismissive.

I wouldn't change definitions like that just because the technology changed, I'm talking about the ability to analyze control flow and logic, not necessarily put code on the screen. What I've seen from most vibe coders is that they don't fully understand what's going on. And I include myself, I tried it for a few months and the code was such garbage after a while that I scrapped it and redid it myself.

Re: A guide to local coding models

#103

Post by SkyPuncher »

simonw wrote: post #4

> I realized I looked at this more from the angle of a hobbiest paying for these coding tools. Someone doing little side projects—not someone in a production setting. I did this because I see a lot of people signing up for $100/mo or $200/mo coding subscriptions for personal projects when they likely don’t need to. Are people really doing that? If that's you, know that you can get a LONG way on the $20/month plans fr…

Time is my limiting factor, especially on personal projects. To me, this makes any multiplying effect valuable.

When I consider it against my other hobbies, $100 is pretty reasonable for a month of supply. That being said, I wouldn’t do it every month. Just the months I need it.

Re: A guide to local coding models

#104

Post by alostpuppy »

simonw wrote: post #50
smcleod wrote: post #30

Earlier quoted context omitted.

On a $20/mo plan doing any sort of agentic coding you'll hit the 5hr window limits in less than 20 minutes.

With Codex it only happened to me once in my 4.5hr session here: https://simonwillison.net/2025/Dec/15/porting-justhtml/ Claude Code is a whole lot less generous though.

For sure. On one project I kept using codex just to see where the wall was. Took a long time.

Re: A guide to local coding models

#105

Post by bitexploder »

jfreds wrote: post #83
lelele wrote: post #75

Earlier quoted context omitted.

> With Sonnet, if you aren't clearing the context very often, you'll hit it within a few hours. Do you mean that users should start a new chat for every new task, to save tokens? Thanks.

Short answer is yes. Not only is it more token-friendly and potentially lower latency, it also prevents weird context issues like forgetting Rules, compacting your conversation and missing relevant details, etc.

Yep. I have Claude snapshot to a markdown doc with key points and resume and iterate. Saves so many tokens.

Re: A guide to local coding models

#106

Post by joshribakoff »

simonw wrote: post #4

> I realized I looked at this more from the angle of a hobbiest paying for these coding tools. Someone doing little side projects—not someone in a production setting. I did this because I see a lot of people signing up for $100/mo or $200/mo coding subscriptions for personal projects when they likely don’t need to. Are people really doing that? If that's you, know that you can get a LONG way on the $20/month plans fr…

To me, it doesn’t matter how cheap open AI codex is because that tool just burns up tokens, trying to switch to the wrong version of node using NVM on my machine. It spirals in a loop and never makes progress, for me, no matter how explicitly or verbosely i prompt.

On the other hand, Claude has been nothing but productive for me.

I’m also confused why you don’t assume people have the intelligence to only upgrade when needed. Isn’t that what we’re all doing? Why would you assume people would immediately sign up for the most expensive plan that they don’t need? I already assumed everyone starts on the lowest plan and quickly runs into session limits and then upgrades.

Also coaching people on which paid plan to sign up for kinda has nothing to do with running a local model, which is what this article is about

Re: A guide to local coding models

#107

Post by tempsaasexample »

bottlepalm wrote: post #101
simonw wrote: post #4

> I realized I looked at this more from the angle of a hobbiest paying for these coding tools. Someone doing little side projects—not someone in a production setting. I did this because I see a lot of people signing up for $100/mo or $200/mo coding subscriptions for personal projects when they likely don’t need to. Are people really doing that? If that's you, know that you can get a LONG way on the $20/month plans fr…

When you pay $1000/month for health insurance and $2000/month for housing.. $200 for something you actually enjoy isn't so bad.

Would you be homeless for 3 days a month so that you could have 30 days of AI?

Not a serious question but I thought it's an interesting way of looking at value.

I used to sell cars in SF. Some people wouldn't negotiate over $50 on a $500 a month lease because their apartment was $4k anyway.

Other people WOULD negotiate over $50 because their apartment was $4k.

Re: A guide to local coding models

#108

Post by m4ck_ »

kristopolous wrote: post #91
simonw wrote: post #4

> I realized I looked at this more from the angle of a hobbiest paying for these coding tools. Someone doing little side projects—not someone in a production setting. I did this because I see a lot of people signing up for $100/mo or $200/mo coding subscriptions for personal projects when they likely don’t need to. Are people really doing that? If that's you, know that you can get a LONG way on the $20/month plans fr…

I use local models + openrouter free ones. My monthly spend on ai models is I'm not cheap, just ahead of the curve. With the collapse in inference cost, everything will be this eventually I'll basically do $ man tool | or even $ cat source | Things I used to do intensively I now do lazily. I've even made a IEITYuan/Yuan-embedding-2.0-en database of my manpages with chroma and then I can just ask my local documentatio…

Is your RAG manpages thing on github somewhere? I was thinking about doing something like that (it's high on my to-do list but I haven't actually done anything with llms yet.)

Re: A guide to local coding models

#109

Post by c-hendricks »

joshribakoff wrote: post #106
simonw wrote: post #4

> I realized I looked at this more from the angle of a hobbiest paying for these coding tools. Someone doing little side projects—not someone in a production setting. I did this because I see a lot of people signing up for $100/mo or $200/mo coding subscriptions for personal projects when they likely don’t need to. Are people really doing that? If that's you, know that you can get a LONG way on the $20/month plans fr…

To me, it doesn’t matter how cheap open AI codex is because that tool just burns up tokens, trying to switch to the wrong version of node using NVM on my machine. It spirals in a loop and never makes progress, for me, no matter how explicitly or verbosely i prompt. On the other hand, Claude has been nothing but productive for me. I’m also confused why you don’t assume people have the intelligence to only upgrade when…

Why is an LLM trying to switch node versions?

Re: A guide to local coding models

#110

Post by deaux »

_delirium wrote: post #78
uneekname wrote: post #63

Earlier quoted context omitted.

Yes, we are doing that. These tools help make my personal projects come to life, and the money is well worth it. I can hit Claude Code limits within an hour, and there's no way I'm giving OpenAI my money.

As a third option, I've found I can do a few hours a day on the $20/mo Google plan. I don't think Gemini is quite as good as Claude for my uses, but it's good enough and you get a lot of tokens for your $20. Make sure to enable the Gemini 3 preview in gemini-cli though (not enabled by default).

Huge caveat: For the $20/mo subscription Google hasn't made clear if they train on your data. Anthropic and OAI on the other hand either clearly state they don't train on paid usage or offer very straightforward opt-outs.

https://geminicli.com/docs/faq/

> What is the privacy policy for using Gemini Code Assist or Gemini CLI if I’ve subscribed to Google AI Pro or Ultra?

> To learn more about your privacy policy and terms of service governed by your subscription, visit Gemini Code Assist: Terms of Service and Privacy Policies.

> https://developers.google.com/gemini-code-assist/resources/p...

The last page only links to generic Google policies. If they didn't train on it, they could've easily said so, which they've done in other cases - e.g. for Google Studio and CLI they clearly say "If you use a billed API key we don't train, else we train". Yet for the Pro and Ultra subscriptions they don't say anything.

This also tracks with the fact that they enormously cripple the Gemini app if you turn off "apps activity" even for paying users.

If any Googlers read this, and you don't train on paying Pro/Ultra, you need to state this clearly somewhere as you've done with other products. Until then the assumption should be that you do train on it.

Post reply on HN