Topic
The practice of retiring older AI model versions, API endpoints, or feature slugs, requiring users to migrate to replacements within a set timeframe.
OpenAI will retire GPT-4o, GPT-4.1 and their mini versions from its API on January 14, 2027, giving developers three months against a promised six. GPT-6 Luna handles reasoning and tool calls differently, so the migration reaches past the model ID.
Reality
Anthropic retires the claude-sonnet-4-5-20250929 model id on November 30, 61 days after its deprecation notice, according to a dev.to walkthrough. Its replacement, claude-sonnet-5-5, costs the same, so the remaining work is finding every key and config that still pins the old string.
Reality
Google priced Nano Banana 2.1 at half Nano Banana 2's per-image rate, $0.0336 for a 1K image, while tripling what it charges for input tokens. Reference-heavy edits can cost more after switching, and no published benchmark backs the quality claims.
Perspective Coverage
10 publishersReality
Google shuts off Nano Banana 2 on the Gemini API on October 29; its replacement, Nano Banana 2.1, lists at about half the price per image. Teams budgeting 4K should plan on the higher of two token counts in Google's own docs until an invoice settles it.
Reality
OpenAI's own deprecations page lists four cutoffs between 24 September and 23 October 2026. Finding out whether they touch your code takes one look at the Usage dashboard, in an account plenty of buyers cannot sign into.
Reality
The retirement covers ChatGPT, ChatGPT Work and Codex on all plans. OpenAI names gpt-5.6-sol as the replacement for Codex with ChatGPT sign-in. The API is exempt, and the inventory of affected tasks lives in one screen.
Reality
A dev.to write-up swaps the single assertion for per-field hit rates with 95 percent intervals over repeated trials. On its own example table, the pair it calls a real regression has intervals that overlap by 0.3 points.
Reality
Reddit users spent late August describing Perplexity answers getting shorter, a Codex allowance draining faster and a Gemini version slipping away, and dev.to says its two-day snapshot of those posts is not a survey.
Reality
The V4.1-Flash change log puts Terminal Bench 2.1 at 90.6 against V4-Pro's 87.9 and DeepSWE at 74.2 against 62.7. DeepSeek says API prices came down with the release and points to a pricing page for the amounts.
Reality
Warner, BMG and Believe are credited as codevelopers of the new models, yet Suno still will not name the catalogs that trained them, and the v5 generation people had tuned their prompts against is switched off.
Reality
The published floor is six months for generally available models, three for Codex and chat variants, and as little as two weeks for anything with preview in the name. The dated entries in the log sit on the floor.
Reality
Free quotas expire on the vendor's schedule, and on AI platforms that schedule is measured in months. The exit costs more than re-plumbing, so it belongs in the integration budget.
Reality