Comment by Aurornis
16 hours ago
For serious work, the difference between spending $10/month and $100/month is not even worth considering for most professional developers. There are exceptions like students and people in very low income countries, but I’m always confused by developers with in careers where six figure salaries are normal who are going cheap on tools.
I find even the SOTA models to be far away from trustworthy for anything beyond throwaway tasks. Supervising a less-than-SOTA model to save $10 to $100 per month is not attractive to me in the least.
I have been experimenting with self hosted models for smaller throwaway tasks a lot. It’s fun, but I’m not going to waste my time with it for the real work.
You need to supervise the model anyway, because you want that code to be long-term maintainable and defect free, and AI is nowhere near strong enough to guarantee that anytime soon. Using the latest Opus for literally everything is just a huge waste of effort.
Yes, but I find supervision much easier and faster with a strong model. It makes fewer dumb mistakes that I have to catch and correct, and it’ll follow my instructions more reliably.
Depends on the task. If it's something that occurs a lot in training data like React/tailwind code then I don't think you need SOTA. Most reasoning models since Sonnet 3.5, Deepseek 3.1 et al will do fine for those tasks.
Doesn't justify 10x the cost in that case imo
Waste of effort... of Opus? If "Opus effort" is cheaper, than dev hours managing yourself more dumb/effective model, what is the point?
rich people dont concern themselves with the cost of tokens.
1 reply →
You don't magically get better results by spending 10x more on a model. If your prompt is crap and harness is crap, you get crap results, regardless of model. And if you run into limits, you aren't working at all.
Buying the most expensive circular saw doesn't get you the best woodworking, but it is the most expensive woodworking.
Not really true. Remember the prompt engineering craze a few years ago with crazy complex prompt composers (langchain) that don’t need to exist any more because the underlying model got so much better at understanding what the humans are actually asking for?
[delayed]
$100 / month will get you rate limited to much to rely on with the Claude plans. People still report getting rate limited on the $200 / plan.
Also not everyone wants to use Claude Code, so if they're paying API pricing it's more likely thousands of dollars a month. If you can get the same results by spending a fraction of that, why wouldn't you?
I got rate limited within an hour on the $200 while working on a single feature.
That was the breaking point, I cancelled my subscription.
As it happens I had a low coding workload over the past two weeks so I've been noodling around in PI mostly with Gemini Flash api. I like it - I even agree it's a much better harness than CC. However, the lock in is real. Even without switching models which each have their own quirks, I expect my work speed to drop drastically for at least a week or two even if I was focused on it fully. But after the learning period I think pi will be faster. The danger of course is that CC is fairly on rails while with PI you could end up spending all your time tinkering with the harness.
And people report getting limited on the $200 plan is putting it very mildly.
You can't do any serious work on it without rationing your work and kneecapping your workflows, to the point where you design workflows around anthropic usage limit woodoo rather than what actually works.
Without this, I run into WEEKLY usage limits on $200 plan, working on a single codebase, one feature at a time, on just day 3.
Thats crazy to pin your entire workflow on. Sorry boss, I can't work today I'm being rate limited by Anthropic :/
For actually serious work, it's a stark difference if your proprietary and security relevant code is sent abroad to a foreign, possibly future hostile country, or is sent to some data center around the corner. It doesn't even need to be defence related.
AFAIK all these companies have SOTA or near-SOTA models available under enterprise licenses. AI companies are not interested in your secret sauce, they are trying to capture the SDLC wholesale.
I’m not sure what you are implying by “enterprise license”, but if you think it provides any meaningful protection against malicious US government actors, you really need to read and internalize the US CLOUD Act.
On a related note, I really need to try some local models (probably starting with qwen), since, at least in 2026, the Chinese models are way better at protecting democracy and free speech than the US models.
If an American company, let's say a company that writes software for power stations, would use the services of a French or Chinese AI company under such enterprise licenses, how long would you think it would take until someone, in Congress e.g., would interfere?
What if they learned that half of the American small and medium sized companies would have started pouring all their business information into such a service?
That doesn't address the concern. Google isn't interested in violating 1st and 4th amendment rights of people who criticize the government... but they do anyway (or more correctly assist the government in doing so).