Upcoming limit, pricing and billing changes
Facts checked on official pages · Subscribe via calendar (.ics)
Deadlines that can break a free plan, bump a bill, or retire a model. Not a ranking. We only list changes with a concrete date from an official page, re-checked on the date above. Confirm on the source before you act.
Upcoming
- GitHub Copilot: selected models deprecated Deprecation GitHub Copilot
Claude Opus 4.8, Claude Opus 5, and GPT-5.6 Terra are deprecated across Copilot experiences on November 16, 2026. Suggested alternatives: Claude Opus 5.5, GPT-6.1 Sol or GPT-6 Luna. Opus 4.8 stays available to individual Copilot subscribers on annual plans.
Official source · verified
- Vercel Hobby: Data Cache migrates to Runtime Cache Billing Vercel
Data Cache is moving to Runtime Cache. Hobby teams migrate on November 29, 2026. Teams created on or after October 30, 2026 already use Runtime Cache by default. Pro teams get a notified migration date. After migration, previously unbilled Data Cache reads/writes are charged at Runtime Cache rates (Hobby keeps free monthly unit allowances).
Official source · verified
- Claude Code: Opus 4.8 fast mode removed Deprecation Claude Code
Fast mode for Opus 4.8 (deprecated as of October 30, 2026) is removed on November 30, 2026. After removal, fast mode requests on Opus 4.8 return an error and do not fall back to standard Opus 4.8. Migrate to Opus 5.5 to keep the speedup.
Official source · verified
- Gemini API: Gemini 4 Argon introductory price ends Pricing Gemini API
Paid-tier Gemini 4 Argon (gemini-4-argon) introductory pricing ends. Official pricing page: input $2.00 → $4.00 and output (including thinking tokens) $10.00 → $20.00 per 1M tokens starting December 1, 2026. Not available on the Free tier.
Official source · verified
- Gemini API: Gemini 3.8 Flash introductory price ends Pricing Gemini API
Paid-tier Gemini 3.8 Flash (gemini-3.8-flash) introductory pricing ends. Official pricing page: input $0.75 → $1.50 and output (including thinking tokens) $3.75 → $7.50 per 1M tokens starting January 1, 2027.
Official source · verified
- Firebase AI Logic: last day to delay App Check auto-enforcement Enforcement Firebase AI Logic
After January 21, 2027 you can no longer delay automatic App Check enforcement for Firebase AI Logic in the Firebase console. Projects that are still delaying will have App Check enforced.
Official source · verified
Recently effective
- Cloudflare D1: 50 MB query result size cap Limit Cloudflare D1
D1 queries whose result exceeds 50 MB fail with “D1 query result exceeded the maximum response size limit of 50 MB”. REST API import and export are exempt. Select needed columns and paginate with LIMIT/OFFSET if you return large results.
Official source · verified
- Firebase AI Logic: App Check auto-enforcement begins Enforcement Firebase AI Logic
Firebase automatically enforces App Check for Firebase AI Logic in projects that were not already enforcing it. Requests without a valid App Check token are blocked, including from released app versions that never implemented App Check.
Official source · verified
Dates and policies change. Each entry links the official page we checked; confirm there before you budget or ship. This page is not a ranking and nobody pays to appear here.
More
AgentRank weekly
Weekly: new AI dev-tool releases, price changes and free-tier catches. Free, unsubscribe anytime.