Capture Official Price Page Drift Before AI API Workbooks Ship

Description and use

An automated recipe for capturing and normalizing AI API pricing data to prevent drift in cost spreadsheets.

Problem

AI API cost workbooks drift faster than teams expect, leading to discrepancies between workbook estimates, SDK examples, and runtime ledgers.

Outcome

Workbooks now reference a verifiable snapshot of pricing data, eliminating drift and providing clear receipts for cost estimates.

Workflow

• 1. Run a snapshot script that fetches the public pricing endpoint (e.g., https://aiwave.live/api/v1/pricing).
• 2. Record metadata: captured_at, pricing_checked, pricing_updated_at, pricing_version, content_sha256, row_count, and the list of models.
• 3. Write the snapshot to pricing‑snapshot.json for version control and diffing.
• 4. Normalize the snapshot to only the fields needed by the workbook (model, provider, base_rate, effective_rate, etc.).
• 5. Import the normalized rows into the workbook, ensuring each row references the snapshot’s captured_at and pricing_version.
• 6. Use the snapshot artifact in release notes and budget reviews to provide a receipt of pricing assumptions.

Evidence

The workbook starts as a useful bridge between engineering, finance, and procurement.

Source

A compact normalized row can look like this:

Source