# Notes

> Dated notes on one real coding task: the USD it cost, and which model did which part. Each note states how many tasks were run. Three tasks are not a ranking.

These notes record a real task, the USD it cost, and which model did which part. They are not a benchmark table, and they are not a ranking of vendors.

Each note says how many tasks were actually run. One task is not a statistic, and three tasks are still three tasks.

Every amount carries the day it was recorded. Published rates move. The card checked for today is on [LLM Pricing](https://bytescope.dev/en/llm-pricing/).

## gemini-3.7-flash-high vs sonnet-5-medium
- Date: 2026-08-14
- [HTML](https://bytescope.dev/en/notes/model-routing-evidence/)
- [Markdown](https://bytescope.dev/en/notes/model-routing-evidence.md)

Three rounds, one task each, on one frozen contract: blind-test scores, Opus review defects, and the 2026-08-14 USD each side spent.