Goal
Read a pricing reference the way the catalogue intends — as a sourced reference, not a live quote — including unit semantics, retrieval date, and freshness state.
Prerequisites
- Read /learn/pricing-references so the references-not-quotes framing is in place.
Step-by-step exercise
Open the pricing hub
Visit /pricing and pick any row with a verified amount.
Expected outcome: You have a specific verified pricing reference to inspect.
Identify the unit and currency
Read the unit column (input vs output tokens, cache reads vs writes, per-second compute) and the currency. Write both down.
Expected outcome: You can name the unit semantics — not just the numeric value.
Open the citation
Click through to the pricing row's citation. Note the retrievedAt date and confirm the URL is the provider's official pricing page.
Expected outcome: You can quote the source URL and the retrievedAt date.
Check the reverification queue for the same source
Visit /reverification?provider=<provider-slug> and confirm whether the same pricing row appears on the queue (stale or review-due).
Expected outcome: You know whether the catalogue currently flags this row as needing re-check.
Completion checklist
Completion checklist
- You named the unit semantics, not just the numeric value.
- You opened the pricing citation directly.
- You checked the reverification queue for the same source.
- You did NOT use the row as a live quote.
Evidence artifact
At the end of this exercise you should have: A short note with provider, unit, retrievedAt, and queue state — copyable into any pricing review document.
Paste the artifact into your design doc, ticket, or PR description. The catalogue's role ends with the artifact; the workload-specific testing is yours.
Related workflow routes
- /select — narrow the source-backed shortlist.
- /compare/build — render verified fields side by side.
- /briefs/build — generate the evidence decision brief.
- /sources — every primary-source citation, by provider.
- /coverage — per-provider verified-field coverage.
- /reverification — sources due for manual re-check.
Exercise does not recommend a model — external testing still required.
Common mistake
Reading a per-token rate as a comparable number across providers. Unit semantics (input vs output, cache reads vs writes, per-second compute) often differ enough that direct numeric comparison misleads.
Example artifact
Illustrative example — not a recommendation. Substitute your own values when you run the workflow.
Example artifact
## Pricing reference review (illustrative)
Provider: <slug>
Source URL: <provider pricing page>
RetrievedAt: <date>
Verbatim from the source:
- Input: <amount> per <unit>
- Output: <amount> per <unit>
Reverification state: <fresh / review-due / stale>
Caveat: Reference only — re-verify before quoting in any contract.Substitute your real shortlist, slugs, dates, and values when you walk the exercise.
Repeat this exercise when
- The reverification queue flags the row as stale.
- The provider publishes a pricing-page update or migration.
- Finance asks for an updated cost projection.
Review before moving on
- Unit semantics are recorded alongside numbers.
- RetrievedAt date is captured.
- Reverification queue state is checked.
- Row is treated as a reference, not a live quote.
Caution: No persistence — checklist resets on every visit. Capture progress in your own notes.
What this exercise does not produce
- A model recommendation. The exercise routes you through evidence — you decide.
- A score or grade for any model. The catalogue does not score.
- A substitute for external prompt, latency, rate-limit, cost, or compliance tests.