Author: Juli Clover
OpenAI today announced the launch of three new GPT models that are available through the OpenAI API. Called GPT-4.1, GPT-4.1 mini, and GPT-4.1 nano, the models are not available for ChatGPT at the current time.
GPT-4.1 includes major improvements in coding, instruction following, and long context, according to OpenAI. The models outperform GPT-4o and GPT-4o mini in all tasks, especially coding tasks. The new models support up to one million tokens of context, and can use that for better long-context comprehension.
On SWE-bench Verified, GPT-4.1 showed a 21.4 percent improvement over GPT-4o for coding, and a 26.6 percent improvement over GPT-4.5. The slimmed down GPT-4.1 mini matches or exceeds GPT-4o with reduced latency and an 83 percent lower cost. GPT-4.1 nano is OpenAI's fastest and cheapest model, ideal for classification, autocompletion, and similar tasks.
OpenAI says that many of the improvements in GPT-4.1 have already been incorporated into the ChatGPT version of GPT-4o, with more functionality to be added in the future. All GPT-4.1 models have a cutoff date of June 2024, which means they are up to date on events that took place before then.
With the introduction of GPT-4.1, OpenAI is deprecating GPT-4.5 in the API, as GPT-4.1 offers much of the same functionality at a lower cost.
GPT-4.1 costs $2 per million input tokens and $8 per million output tokens, while GPT-4.1 mini costs $0.40 per million input tokens and $1.60 per million output tokens. GPT-4.1 nano is just $0.10 per million input tokens and $0.40 per million output tokens. Customized models are available at a higher price.
Tag: OpenAI
This article, "OpenAI Launches New Coding-Focused GPT-4.1 Models" first appeared on MacRumors.com
Discuss this article in our forums
Continue reading...
OpenAI today announced the launch of three new GPT models that are available through the OpenAI API. Called GPT-4.1, GPT-4.1 mini, and GPT-4.1 nano, the models are not available for ChatGPT at the current time.
GPT-4.1 includes major improvements in coding, instruction following, and long context, according to OpenAI. The models outperform GPT-4o and GPT-4o mini in all tasks, especially coding tasks. The new models support up to one million tokens of context, and can use that for better long-context comprehension.
On SWE-bench Verified, GPT-4.1 showed a 21.4 percent improvement over GPT-4o for coding, and a 26.6 percent improvement over GPT-4.5. The slimmed down GPT-4.1 mini matches or exceeds GPT-4o with reduced latency and an 83 percent lower cost. GPT-4.1 nano is OpenAI's fastest and cheapest model, ideal for classification, autocompletion, and similar tasks.
OpenAI says that many of the improvements in GPT-4.1 have already been incorporated into the ChatGPT version of GPT-4o, with more functionality to be added in the future. All GPT-4.1 models have a cutoff date of June 2024, which means they are up to date on events that took place before then.
With the introduction of GPT-4.1, OpenAI is deprecating GPT-4.5 in the API, as GPT-4.1 offers much of the same functionality at a lower cost.
GPT-4.1 costs $2 per million input tokens and $8 per million output tokens, while GPT-4.1 mini costs $0.40 per million input tokens and $1.60 per million output tokens. GPT-4.1 nano is just $0.10 per million input tokens and $0.40 per million output tokens. Customized models are available at a higher price.
Tag: OpenAI
This article, "OpenAI Launches New Coding-Focused GPT-4.1 Models" first appeared on MacRumors.com
Discuss this article in our forums
Continue reading...