| This text initially appeared on Medium. Tim O’Brien has given us permission to repost right here on Radar. |
If you’re working with AI instruments like Cursor or GitHub Copilot, the true energy isn’t simply gaining access to totally different fashions—it’s figuring out when to make use of them. Some jobs are OK with Auto. Others want a stronger mannequin. And generally you need to bail and change in case you proceed spending cash on a fancy drawback with a lower-quality mannequin. Should you don’t, you’ll waste each money and time.
And that is the lacking dialogue in code technology. There are just a few “camps” right here; nearly all of folks writing about this seem to view this as a fantastical and enjoyable “vibe coding” expertise, and some folks on the market try to make use of this expertise to ship actual merchandise. If you’re in that final class, you’ve most likely began to appreciate that you could spend a implausible sum of money in case you don’t have a technique for mannequin choice.
Let’s make it very particular—in case you join Cursor and drop $20/month on a subscription utilizing Auto and you’re proud of the output, there’s not a lot to fret about. However if you’re beginning to run brokers in parallel and are paying for token consumption atop a month-to-month subscription, this put up will make sense. In my very own expertise, a single developer working alone can simply spend $200–$300/day (or 4 instances that determine) if they’re attempting to deal with a mission and have opted for the most costly mannequin.
And, if you’re an organization and also you give your builders limitless entry to those instruments. Prepare for some surprises.
My Escalation Ladder for Fashions…
- Begin right here: Auto. Let Cursor path to a robust mannequin with good capability. If output high quality degrades or the loop happens, escalate the difficulty. (Cursor explicitly says Auto selects amongst premium fashions and can change when output is degraded.)
- Medium-complexity duties: Sonnet 4/GPT‑5/Gemini. Use for targeted duties on a handful of information: strong unit checks, focused refactors, API remodels.
- Heavy raise: Sonnet 4 – 1 million. If I must do one thing that requires extra context, however I nonetheless don’t need to pay prime greenback, I’ve been beginning to transfer up fashions that don’t shortly max out on context.
- Ultraheavy raise: Opus 4/4.1. Use this out when the duty spans a number of initiatives or requires lengthy context and cautious reasoning, then change again as soon as the large transfer is completed. (Anthropic positions Opus 4 as a deep‑reasoning, lengthy‑horizon mannequin for coding and agent workflows.)
Auto works fantastic, however there are occasions when you may sense that it’s chosen the unsuitable mannequin, and in case you use these fashions sufficient, you realize when you find yourself taking a look at Gemini Professional output by the verbosity or the ChatGPT fashions by the best way they go about fixing an issue.
I’ll admit that my heavy and ultraheavy decisions listed here are biased in direction of the fashions I’ve had extra expertise with—your individual expertise would possibly range. Nonetheless, you must also have the same escalation checklist. Begin with Auto and solely improve if it’s worthwhile to; in any other case, you will be taught some classes about how a lot this prices.
Watch Out for “Pondering” Mannequin Prices
Some fashions assist express “pondering” (longer reasoning). Helpful, however costlier. Cursor’s docs observe that enabling pondering on particular Sonnet variations can rely as two requests underneath group request accounting, and within the particular person plans, the identical concept interprets to extra tokens burned. Briefly, pondering mode is great—use it whenever you want it.
And when do you want it? My rule of thumb right here is that once I perceive what must be executed already, once I’m asking for a unit check to be polished or a way to be executed within the sample of one other… I normally don’t want a pondering mannequin. However, if I’m asking it to research an issue and suggest varied choices for me to select from, or (one thing I do typically) once I’m asking it to problem my choices and play satan’s advocate, I’ll pay the premium for one of the best mannequin.
Max Mode and When to Use It
Should you want large context home windows or prolonged reasoning (e.g., sweeping modifications throughout 20+ information), Max Mode may help—however it’ll eat extra utilization. Make Max Mode a non permanent software, not your default. If you end up continuously requiring Max Mode to be turned on, there’s a great likelihood you’re “overapplying” this expertise.
If it must eat one million tokens for hours on finish? That’s normally a touch that you just want one other programmer. Extra on that later, however what I’ve seen too typically are managers who assume that is just like the “vibe coding” they’re witnessing. Spoiler alert: Vibe coding is that factor that individuals do in shows as a result of it takes 5 minutes to make a foolish online game. It’s 100% not programming, and to make use of codegen, right here’s the key: You must perceive how you can program.
Max Mode and pondering fashions are usually not a shortcut, and neither are they a alternative for good programmers. Should you assume they’re, you will be paying prime greenback for code that can at some point should be rewritten by a great programmer utilizing these similar instruments.
Most Necessary Tip: Watch Your Invoice as It Occurs
A very powerful tip is to recurrently monitor your utilization and utilization charges in Cursor, since they seem inside a minute or two of working one thing. You possibly can see utilization by the minute, the variety of tokens consumed, and in some circumstances, how a lot you’re being charged past your subscription. Make a behavior of checking a few instances a day, particularly throughout heavy classes, and ideally each half hour. This helps you catch runaway prices—like spending $100 an hour—earlier than they get out of hand, which is solely potential in case you’re working many parallel brokers or doing resource-intensive work. Paying consideration ensures you keep in charge of each your utilization and your invoice.
Maintain Observe and Keep away from Loops
The opposite factor it’s worthwhile to do is preserve monitor of what works and what doesn’t. Over time, you’ll discover it’s very straightforward to make errors, and the fashions themselves can generally fall into loops. You would possibly give an instruction, and as a substitute of resolving it, the system retains working the identical course of many times. Should you’re not paying consideration, you may burn via numerous tokens—and some huge cash—with out really getting sound output. That’s why it’s important to observe your classes intently and be able to interrupt if one thing appears prefer it’s caught.
One other pitfall is pushing the fashions past their limits. There are duties they’ll’t deal with nicely, and when that occurs, it’s tempting to maintain rephrasing the request and asking once more, hoping for a greater consequence. In apply, that always results in the identical cycle of failure, besides you’re footing the invoice for each try. Figuring out the place the boundaries are and when to cease is crucial.
A sensible method to keep on prime of that is to keep up a working diary of what labored and what didn’t. Report prompts, outcomes, and notes about effectivity so you may be taught from expertise as a substitute of repeating costly errors. Mixed with keeping track of your stay utilization metrics, this behavior will assist you refine your method and keep away from losing each money and time.
