GPT-4.1

Overview

OpenAI's April 2025 coding-focused model, since superseded by the GPT-5 line (GPT-5.6 is current). GPT-4.1 offered a 1M token context window and marked improvements in structured output and complex task completion.

Why it matters

GPT-4.1 was OpenAI's response to coding-centric competitive pressure from Claude Sonnet and DeepSeek, and it set the template — 1M context, strong instruction-following, aggressive pricing — that the GPT-5 family later refined. It remains available and inexpensive, but new builds should evaluate the GPT-5.6 tiers first: OpenAI's release cadence means each GPT-4-era model is now a legacy choice, kept alive mainly for teams with pinned integrations that have not yet scheduled a migration.