Reduce what a review spends without losing the findings you care about. The levers are ordered by money saved per unit of effort: the first two are configuration changes that cost you no coverage, the later ones trade coverage or convenience for spend.
Keep two figures apart - what your provider bills you, and what ProPR reports you spent. The second is only as good as the prices it has, which is why pricing comes before any tuning.
1. Choose a provider family that caches
Prompt caching is the lever worth the most money, and the connection’s provider family decides whether you get it - so it is settled when you create the connection rather than tuned afterwards. Which families cache, and what a gateway buys you in return, is in prompt caching.
If you reach a cacheable model through a gateway today, add a second connection on its native family and repoint the logical model at it. No client review configuration changes, because nothing in the review loop names a provider - see how model selection works. Keep the gateway connection active alongside the new one until you have compared two reviews.
2. Make the prices right before you tune anything
Confirm your negotiated rates are entered, and that no model in use is reporting no price at all. Do this first; every decision below is judged on these numbers. See pricing overrides, and for a model with no price, the per-edition remedies.
3. Spend reasoning effort where it changes the answer
Effort is a property of the logical-model name - see reasoning effort - which is what lets you spend it unevenly. Where it is worth spending:
- The name mapped to Review default is the expensive one: it applies to every file of every review.
- Triage and Verification are the cheap half of the job. Point them at a smaller model and an ordinary file costs less without touching the hard ones.
- Per-tier names go further - a small model on Low effort, the expensive one only on High effort.
All of them are mapped on one screen - see purposes.
Baseline reasoning effort is a separate per-client setting whose default leaves cost unchanged. Leave it there unless you have measured a reason not to - see what you can tune.
4. Count your passes
The review pass list is where a small change multiplies. Read the client’s list and its multi-pass union switch, and justify each entry - see review passes.
Two habits follow. A shadow pass earns its cost while you are evaluating a model or a lens and not afterwards, so remove it when the evaluation is over. Evidence-backed verification trades the other way, buying back correct findings with extra model calls, so decide it deliberately rather than leaving it as found - see what you can tune.
5. Review less
- Exclusions. Generated, vendored and mechanical files are the cheapest tokens to save, because they are never sent. The patterns live in the repository rather than in ProPR - see configuring ProPR from your repository.
- Incremental re-reviews. Already the default, and need no configuration - see what happens during a review.
- Iteration budgets. They bound how long one file’s review may keep gathering context before it must conclude - see review loop budgets.
What does not save money: the minimum severity to post, and disabling SCM comment posting. Both act after every model call has been made and paid for. They are publication controls - see run reviews without posting them.
6. Cap what a client can spend
Budget caps are the backstop rather than a tuning lever: they stop spend, they do not reduce it. Read what you can tune before you set one, because the scopes do not behave alike. Budgeting needs a commercial license - see editions.
If you run ProCursor, its indexing and querying spend embedding tokens reported separately from review usage. Read what it costs before widening a source’s root path or adding tracked branches.
Confirm it worked
- Run two reviews of comparable size, before and after the change, and compare the per-client token and spend reporting.
- Open the job protocol on each. It lists every pass and every model call, which is the fastest way to find a pass you forgot to remove or an effort setting applying more widely than you thought - see review diagnostics.
- If a model still shows no price, go back to step 2 above: the figures are wrong, not the spend.
For a symptom this page does not cover, troubleshooting routes it.