Using GPT-4 Turbo? It stops responding July 1 — migrate to GPT-4.1

OpenAI logoOpenAIAct nowJune 8, 2026Deprecation
What happened
OpenAI set a hard shutdown date for GPT-4 Turbo: all endpoints stop responding July 1, 2026.
Why it matters
Anything still pinning `gpt-4-turbo` — code or automation workflows — fails silently on that date.
What to do
Switch to GPT-4.1 now and convert legacy function-calling payloads to the tools format.

OpenAI has set a hard deprecation date for GPT-4 Turbo: all endpoints stop responding on July 1, 2026. Unlike previous soft deprecations, there is no extended sunset period for existing API keys.

For most chat and completion workloads, GPT-4.1 is a drop-in replacement — same request shape, comparable latency, lower per-token cost. The exception is function calling: GPT-4.1 rejects the legacy function_call parameter and requires the tools array format.

If you run automations through n8n, Make, or Zapier, check each OpenAI node — older workflow templates still pin gpt-4-turbo explicitly and will start failing silently on July 1.

What to do

  1. 1 Inventory every workflow and codebase reference pinning `gpt-4-turbo`
  2. 2 Switch model id to `gpt-4.1` and convert legacy `function_call` payloads to the `tools` format
  3. 3 Re-run critical automations once before the deadline — do not wait for July 1 failures

Affected tools & models

OpenAI APIn8nZapier

Never need to catch up again

The weekly delta — only verdict changes and act-now items. No digest filler.