Skip to content
Nilay Kabariya

Home · Topic · 5 entries

AI model retirements: OpenAI, Claude and Gemini errors, tested

AI providers retire models on a schedule, and code that worked for months starts failing the day after: a 404 for a model name that is gone, a 400 for a parameter the replacement no longer accepts, or a quota error that names a different model than the one you called. Every entry below was reproduced against the real API, with the exact error, the retirement dates from the provider, and what to change.

Entries

  1. No.060

    Claude "temperature is deprecated for this model" (400)

    Switching to Claude Sonnet 5.5, Haiku 5.5 or Opus 4.7 breaks code that sets temperature, even temperature: 0. Every case tested, in Node and Python.

    > BadRequestError: 400 {"type":"error","error":{"type":"invalid_request_error","message":"`temperature` is deprecated for this model."}}

    AI4 min
  2. No.059

    OpenAI Oct 23 model shutdown: the 404, list, replacements

    gpt-3.5-turbo, gpt-4, o1, o3-mini and more stopped working on October 23, 2026. The real error, tested after the shutdown, and a check that misleads.

    > NotFoundError: 404 The model `gpt-3.5-turbo` has been deprecated, learn more here: https://platform.openai.com/docs/deprecations

    AI4 min
  3. No.049

    Gemini 2.5 Flash "no longer available to new users": the fix (404)

    New Gemini API keys get a 404 on 2.5 Flash and Pro. Which model names answer instead, tested, and why December 1 quietly moves old keys to Flash-Lite.

    > {

    AI5 min
  4. No.048

    Gemini API 429 RESOURCE_EXHAUSTED: why your limit is 0

    Gemini's 429 hides three different limits, and one of them never resets. Each hit on purpose with a free key, plus the 503 high demand error.

    > {

    AI5 min
  5. No.044

    Groq: llama-3.3-70b-versatile does not exist, 413 and 429

    Groq says the model llama-3.3-70b-versatile does not exist? That and its look-alike 413 and 429 errors, reproduced on the free tier with the fix for each.

    > 413 Request too large for model `openai/gpt-oss-20b` in organization `org_…` service tier `on_demand` on tokens per minute (TPM): Limit 8000, Requested 30074, please reduce your message size and try again.

    AI4 min

Post card · Every Sunday

What broke this week.

One email on Sundays: the Node, TypeScript, Python and AI SDK releases that broke builds that week, each one reproduced, with a link to the fix that worked. Never shared, and you can leave any time.

— Nilay

or follow by RSSor on X

By subscribing you agree to the privacy note. One click to leave.

tip: paste the exact error text