docs(data): Oct 2026 official Ultrafast and Gemini model lifecycle update

[blackkcold]

Officially verified material changes

• OpenAI GPT-6.1 Sol Ultrafast launched Oct 8: API $12/M input and $60/M output (6× Standard); eligible Pro $500 / Enterprise / Edu Work and Codex; Pro $500 allowance 8× and credit 6×. Plus / Pro 100 / Pro 200 / Business not eligible.
• Google Gemini API Oct 8: Gemini 3.7 Flash and 3.5 Flash deprecated / redirected; legacy Deep Research Agent preview shutdown scheduled Oct 23.

Scope

• Refresh data.js plan metadata, official sources and snapshot date; visible website policy note and README.
• Keep AA measurement points, rankings, model/Plan recommendations and long Agent correction unchanged. No fabricated benchmark point for a speed tier.
• Preserve MIT, CODEOWNERS, CI and Pages.

Preflight validation

• V8 syntax parse of app.js and modified data.js passed, equivalent to the repository's Node syntax-check scope.
• Required files confirmed present: index.html, styles.css, app.js, data.js, README.md, LICENSE.

Official sources

• https://developers.openai.com/api/docs/models/gpt-6.1-sol
• https://help.openai.com/en/articles/20001516-managing-usage-with-gpt-6-astra-in-work-and-codex
• https://ai.google.dev/gemini-api/docs/changelog

Await GitHub Static validation and required review before merging.