GPT-4.1 shipped to the API with a million-token context window — but not to ChatGPT, which is where most reps actually work.
A million-token window means a rep could in principle drop an entire account history into one session. In practice this landed in the API first, so unless your team runs its own tooling, nothing changed on your desktop that day.
OpenAI released three models — GPT-4.1, GPT-4.1 mini and GPT-4.1 nano — on 14 April 2025. All three carry a context window of up to one million tokens and a knowledge cutoff of June 2024. GPT-4.1 scored 54.6% on SWE-bench Verified and 38.3% on Scale's MultiChallenge instruction-following benchmark.
The models launched in the API only. They were not available in ChatGPT on announcement day.
Two minutes, once a week. What changed in AI, and what to run because of it.