GPT-4.1 Launches with Major Enhancements

OpenAI Launches GPT-4.1 Series with Enhanced Capabilities

• OpenAI has released the GPT-4.1 series, including mini and nano versions, with major improvements in coding, instruction following, and context understanding.
• The flagship GPT-4.1 model scores 54.6% on SWE-bench Verified for coding tasks, a 21.4-point improvement over GPT-4o.
• GPT-4.1 outperforms previous models by 10.5 points on the MultiChallenge benchmark for instruction following.
• The new models can handle up to 1 million tokens, significantly increasing from the previous generation’s 128,000 tokens.
• GPT‑4.1 is priced 26% lower than GPT‑4o for median queries, with the nano version being the fastest and most affordable.

The newly launched GPT-4.1 series by OpenAI offers significant advancements in performance metrics such as coding tasks and instruction following while reducing costs by up to 26% compared to earlier versions.