We’re at around $100/mo/engineer, mostly for Cursor. We’re using it for augmentation, not vibing, and review every line (with less detailed attention to tests), because we’re building the foundation of something that needs to be secure and stable for several years.
The Microsoft GitHub Copilot Pro Plus Premium Ultra Max at $40 a month is enough for me. My job isn’t only about coding.
Moreover, if you spend that much on tokens, that sounds like a skill issue and you may be creating a lot of technical debt. I don’t see how anyone can have the brain capacity to handle enormous code bases.
~$250 in 3 weeks at work. Its not steady, equal token distribution per day. There are days when it is $40 and there are days when it is < $10. Mostly due to bottlenecks like waiting on reviews, figma designs, story refinement, etc.
(Sidebar: There is a prediction that the traditional roles of Designer, Product Owner and Programmer are disappearing and converging into one single specialized role. (Claude Code has a blog about this) and I feel there is truth in this.)
So, my runrate right now is ~$4,200 per year, but I won't be surprised if it goes up. It depends on several factors.
It is crazy when I hear how much some people spend on AI! Having an agent cut out some of the boilerplate code is useful, but beyond that, are people just creating a backlog wishlist and then walking away?
That’s exactly what we’re doing. Connect the agent to your GitHub issues, tell it implement each one and spin on a loop until it’s finished. There’s more nuance to it than that but at a high level yeah, that’s how some people are using it.
$40/month at work and $10/month at home, and it's more than I can use.
I cannot imagine productively spending $250k/year on LLM coding - you'd need some kind of massive tree of agents reviewing each other's work and I think even then you would struggle to keep them on-task and sanity-checked. However, I don't make $500k a year so what do I know...
This is the right question but hard to answer in practice. Shipped features vary too much in complexity to compare directly. What you can actually track is cycle time: time from ticket opened to merged PR. That's gone from ~3 days to ~6 hours on greenfield work for us.
The catch is that token spend and quality aren't correlated the way you'd expect. Low-spend months when I'm directing carefully and reviewing every diff tend to produce better code than high-spend months where I'm letting agents run longer chains. The expensive runs generate more code, not necessarily better code.
Jensen's $250k figure only makes sense if you're running dozens of parallel agents continuously. Most engineers are doing something more like augmented pairing. The unit economics are actually pretty good at $100-200/month per person. Beyond that you're hitting diminishing returns unless you've built actual agent infrastructure to parallelize and verify the work.
We’re at around $100/mo/engineer, mostly for Cursor. We’re using it for augmentation, not vibing, and review every line (with less detailed attention to tests), because we’re building the foundation of something that needs to be secure and stable for several years.
The Microsoft GitHub Copilot Pro Plus Premium Ultra Max at $40 a month is enough for me. My job isn’t only about coding.
Moreover, if you spend that much on tokens, that sounds like a skill issue and you may be creating a lot of technical debt. I don’t see how anyone can have the brain capacity to handle enormous code bases.
~$250 in 3 weeks at work. Its not steady, equal token distribution per day. There are days when it is $40 and there are days when it is < $10. Mostly due to bottlenecks like waiting on reviews, figma designs, story refinement, etc.
(Sidebar: There is a prediction that the traditional roles of Designer, Product Owner and Programmer are disappearing and converging into one single specialized role. (Claude Code has a blog about this) and I feel there is truth in this.)
So, my runrate right now is ~$4,200 per year, but I won't be surprised if it goes up. It depends on several factors.
It is crazy when I hear how much some people spend on AI! Having an agent cut out some of the boilerplate code is useful, but beyond that, are people just creating a backlog wishlist and then walking away?
That’s exactly what we’re doing. Connect the agent to your GitHub issues, tell it implement each one and spin on a loop until it’s finished. There’s more nuance to it than that but at a high level yeah, that’s how some people are using it.
We're sitting around $300/month across a couple of tools. Felt like a lot initially, then we know it's actually worth it
$40/month at work and $10/month at home, and it's more than I can use.
I cannot imagine productively spending $250k/year on LLM coding - you'd need some kind of massive tree of agents reviewing each other's work and I think even then you would struggle to keep them on-task and sanity-checked. However, I don't make $500k a year so what do I know...
You know, wouldn't it be better to ask, how much is the avegare cost of shipped feature and if that has gone down at all?
This is the right question but hard to answer in practice. Shipped features vary too much in complexity to compare directly. What you can actually track is cycle time: time from ticket opened to merged PR. That's gone from ~3 days to ~6 hours on greenfield work for us.
The catch is that token spend and quality aren't correlated the way you'd expect. Low-spend months when I'm directing carefully and reviewing every diff tend to produce better code than high-spend months where I'm letting agents run longer chains. The expensive runs generate more code, not necessarily better code.
Jensen's $250k figure only makes sense if you're running dozens of parallel agents continuously. Most engineers are doing something more like augmented pairing. The unit economics are actually pretty good at $100-200/month per person. Beyond that you're hitting diminishing returns unless you've built actual agent infrastructure to parallelize and verify the work.
We are way past this question now.
About $50 a day