GPT-4.1 nano is a compact large language model developed by OpenAI as part of the GPT-4.1 series released on April 14, 2025. It is positioned as the most lightweight, cost-efficient, and fastest variant in the 4.1 family, specifically optimized for developers requiring low-latency performance and on-device AI applications. The model is designed for high-frequency tasks such as text classification, real-time autocompletion, and metadata extraction from large-scale datasets.
One of the model's defining features is its 1 million token context window, a significant expansion from previous lightweight generations. This capacity allows GPT-4.1 nano to process entire code repositories or voluminous documents in a single request. Despite its reduced size, the model demonstrates improved reasoning and coding capabilities over its predecessor, GPT-4o mini, and maintains multimodal support for both text and image inputs. The model has a knowledge cutoff of June 2024 and supports supervised fine-tuning to allow for domain-specific optimization.