GPT-4.1

GPT-4.1 is OpenAI’s April 2025 model, released as an API-only update aimed squarely at developers rather than ChatGPT users. Its headline feature is a 1-million-token context window across all three variants, GPT-4.1, GPT-4.1 Mini, and GPT-4.1 Nano, an eightfold jump over GPT-4o’s 128,000 tokens. It scores 80.1% on MMLU and 54.6% on SWE-bench Verified, a 21-point improvement over GPT-4o on real-world software engineering tasks pulled from GitHub issues.

OpenAI priced the flagship variant at $2 per million input tokens and $8 per million output tokens, about 26% cheaper than GPT-4o, while GPT-4.1 Mini undercuts GPT-4o by roughly 83% on many benchmarks. The combination of long context and coding performance made it popular quickly for tools that need to process large codebases or long documents in a single pass, and it effectively replaced GPT-4.5 as OpenAI’s recommended model for API use later in 2025.