5 days Ago By Kahekashan
OpenAI has officially introduced its latest AI model series, GPT-4.1, pushingthe boundaries of what generative AI can achieve—especially in the world ofsoftware development. This release includes three new variants: GPT-4.
1,GPT-4.1 mini, and GPT-4.1 nano, each designed to deliver improved codingcapabilities and better context understanding.
In arecent blog post, OpenAI highlighted key features of the GPT-4.1 family. “Theyalso have larger context windows—supporting up to 1 million tokens ofcontext—and are able to better use that context with improved long-contextcomprehension.
They feature a refreshed knowledge cutoff of June 2024.” Theseupdates are part of the industry-wide movement to build AI tools capable ofhandling increasingly complex programming tasks. Thislaunch comes as competition in the AI space heats up.
Google recently unveiledGemini 2.5 Pro, a model that also boasts a 1-million-token context window andstrong performance on various coding benchmarks. Meanwhile, Anthropic’s Claude3.
7 Sonnet and China-based DeepSeek’s V3 model are also gaining momentum amongdevelopers and tech communities. Accordingto OpenAI, GPT-4.1 delivers a notable upgrade over its predecessor, GPT-4o.
Itexcels in a range of programming tasks, from solving problems autonomously tohandling frontend coding more effectively. GPT-4.1 also minimizes unnecessaryedits, follows diverse formats more accurately, and ensures consistent use oftools across workflows.
As a result, OpenAI has announced it will phase out theGPT-4.5 Preview from its API. Theretirement of GPT-4.
5 Preview is scheduled for July 14, 2025, giving developersenough time to make the switch. Originally released as a research preview toexplore the potential of high-performance AI models, GPT-4.5 served as avaluable experiment.
OpenAI says that the positive qualities from GPT-4.5—suchas creativity, writing nuance, humor, and style—will be integrated into futuremodels. All threeGPT-4.
1 models are now available to developers. “Through efficiencyimprovements to our inference systems, we’ve been able to offer lower prices onthe GPT 4.1 series,” OpenAI shared.
GPT-4.1 is reportedly 26% cheaper thanGPT-4o for average queries, while GPT-4.1 nano is described as the “cheapestand fastest model ever.
” Pricing Details: GPT-4.1 costs $2 per million input tokens and $8 per million output tokens. GPT-4.
1 mini is priced at $0.40/million input tokens and $1.60/million output tokens.
GPT-4.1 nano comes in at just $0.10/million input tokens and $0.
40/million output tokens. OpenAIhas also enhanced its discount for repeated queries by increasing the promptcaching discount from 50% to 75% for these models. Additionally, long-contextqueries will now be supported at no extra charge beyond the standard per-tokenrates.
With thislatest release, OpenAI continues to refine its offerings, making advanced AImore accessible, affordable, and capable of meeting the demands of modernsoftware development..
