feat: add model-specific cache pricing

[pjb157]

Summary

• normalize model-specific prompt-cache pricing from the admin models response
• add calculated cache-read prices to the model catalog and supported model pages
• replace the generated catalog’s blanket discount claim with model-specific guidance

Impact

Customers can see the current cache-read price per 1M input tokens for each supported model. Unsupported models display an unavailable marker, and model detail pages only show cache pricing when the required metadata and realtime tariff are present.

Validation

• npm run test:run -- src/lib/models.test.ts src/lib/model-artifacts.test.ts (10 tests pass)
• scoped ESLint on the four changed files
• npm run build
• git diff --check

The full test suite retains one unrelated existing failure in src/lib/search.test.ts (stripMarkdown removes inline code instead of preserving its text).

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
documentation	Ready	Preview	Jul 31, 2026 11:34am

[Copilot]

Pull request overview

This PR adds support for model-specific prompt-cache pricing metadata to the docs site’s model pipeline, and surfaces a calculated cache-read price (per 1M input tokens) in both the models catalog page and individual model pages when the required realtime tariff + cache metadata are present.

Changes:

• Extend model normalization (transformModels) to parse cache_pricing into a typed cachePricing field.
• Compute and expose cacheReadPricePer1M (+ multiplier) in generated model artifacts, and render it in the models index table + model detail markdown.
• Update copy in the models index markdown to remove the blanket “~10%” cache discount claim in favor of model-specific guidance.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated no comments.

File	Description
src/lib/models.ts	Adds ModelCachePricing, parses cache_pricing from the admin models response, and exports transformModels for direct testing.
src/lib/models.test.ts	Adds focused tests covering cache pricing normalization (enabled/disabled/absent).
src/lib/model-artifacts.ts	Calculates cache-read pricing when metadata + realtime pricing exist, renders it in the index table and model markdown, and factors out renderModelsIndexMarkdown for testing.
src/lib/model-artifacts.test.ts	Adds tests ensuring cache-read pricing is only computed when appropriate and is rendered (or omitted) correctly in markdown outputs.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.