Picture this: last week you used Opus-4.8 and tasked it with a rename-and-wire-up job. It did this job perfectly, because of course it did. However, you paid roughly 5x more than you needed to for this. Your default model is overpowered compared to what’s available now. You’re taking the Ferrari to pick up the groceries.

On June 30th, Anthropic released Claude Sonnet 5, which is almost as good as Opus-4.8, but much cheaper. They also made it the default model in Claude Code for Free and Pro plans (if you’re on these last two, then this newsletter won’t really mean much to you, save yourself a few minutes and move on).

If you’re on the paid API - and to a lesser extent the Max plans - then this subject is of interest to you, and you might have missed it.

You need a system to select the model your prompt invokes.

Sonnet 5 is, by Anthropic’s own measurements, almost as good in every way as Opus-4.8:

Except in one important metric: Sonnet 5 costs somewhat less than Opus 4.8. Sonnet 5 is $2/$10 Mtok vs $5/$25 Mtok on Opus.

In other words, the gap you’re paying extra for just got thin. There’s not much air now between Sonnet 5 and Opus 4.8.

Thus far you’ve probably decided that benchmarking models every two days is counter-productive and that you can move a lot faster by just choosing one model and focusing on your output. It’s not a dumb approach.

However, at scale this approach will cost you millions. Every task you send to the frontier model out of habit is now paying the Opus Tax.

It's time to start being a little more selective. Opus 4.8 still wins at higher effort tasks (it may even be more cost effective at these) according to HN consensus. Model selection is no longer pedantic. It's resource allocation. You need to size it like a server.

We use a four-axis heuristic for model selection, with each scoring 1-3 points:

Scoring based on this leads to the following partition of tasks (in the context of Claude Code models)

In addition, any risk score of 3 bumps you up one tier automatically.

Here’s a worked example: this week at Temrel we were tasked with adding a new admin-only form to the Maitris backend. Relatively simple stuff. Here was our decision.

Note the bump rule never fired here. If that same form had needed a destructive migration on production data, Risk goes to 3 and the whole task bumps up a tier, regardless of the total.

Sonnet 5 overthinks smaller tasks, thinking longer, using more tokens and generally doing more work than is asked for, or necessary. In the aggregate, this will eat into your productivity and cost effectiveness significantly.

The lesson is that we need to be systematically more selective with our models. Scoping tasks ahead of prompt execution is part of cost control, and therefore part of context engineering.

Effort is now a parameter, set with words.

This week’s free Temrel tool is the model-router. Shipping the four-axis heuristic above as a Claude Code skill, plus a CLI that parses local Claude Code transcripts and reports the actual cost-per-task by model.

The model-router audit function in this toolset flags where you paid Opus prices for work Sonnet’s profile covers and estimates the total overspend. Run it on your own machine and check the last 30 days.

Caveat: the exact dollar figures depend on a pricing file that you must verify and Claude Code’s 30-day rolling transcript window caps how far back you can go.

Model routing is about to become a default, must-have skill, just like instance sizing. As agents multiply, per-task allocation compounds really quickly into both speed and money.

As an agentic engineer, your job description just grew by one line: resource allocation is part of the craft now.

What's your default model, and when did you last actually reconsider it?

Send me your tier split after a week of routing; I'll share the distribution in a future issue.

See you next Thursday.

Ben