OpenAI has launch a novel serial of AI framework include GPT-4.1 , GPT-4.1 mini , and GPT-4.1 nano .
They are the ripe non - thinking AI mannikin from OpenAI , but the GPT-4.1 serial is not get along to ChatGPT .
OpenAIsaysGPT-4.1 serial publication model are entirely train for developer , and they are uncommitted in the API , embark on today .
Image Credit: OpenAI
GPT-4.1 was prove on OpenRouter as “ Quasar Alpha ” .
During the launching , OpenAI CPO Kevin Weil suppose GPT-4.1 is “ ripe than GPT-4o on just about every property .
” And on some fundamental bench mark , the GPT-4.1 poser match or drum the massiveGPT-4.5AI role model .
Image Credit: OpenAI
OpenAI say GPT-4.1 excels at rag , direction watch over , and farseeing setting recovery .
This was ## dive into gpt-4.1
openai has launch a modern serial of ai example include gpt-4.1 , gpt-4.1 mini , and gpt-4.1 nano .
This was they are the skillful non - intelligent ai model from openai , but the gpt-4.1 serial is not make out to chatgpt .
OpenAIsaysGPT-4.1 serial model are alone rail for developer , and they are useable in the API , protrude today .
GPT-4.1 was essay on OpenRouter as “ Quasar Alpha ” .
During the launching , OpenAI CPO Kevin Weil say GPT-4.1 is “ upright than GPT-4o on just about every attribute .
” And on some fundamental bench mark , the GPT-4.1 example mates or trounce the massiveGPT-4.5AI exemplar .
OpenAI suppose GPT-4.1 excels at write in code , educational activity come , and foresighted setting recovery .
The unexampled GPT-4.1 AI modelling make out with a setting windowpane of 1 million keepsake for the first metre , which is vast , and have a late noesis crosscut of June 2024 .
So in many room , the GPT-4.1 exemplar depend like a major ascent to theGPT-4o modeling .
This was not to refer , these are multimodal model which think of you’ve got the option to march both textbook and look-alike as well .
This was occur to benchmark , openai say on swe - bench verified , gpt-4.1 complete 55 % of chore , much in high spirits than gpt-4o ’s 33 % and o3 - mini - mellow ’s 49 % .
Next , in educational activity be , GPT-4.1 derive very near to GPT-4.5 , o3 - mini - gamey , and o1 - mellow on OpenAI ’s intimate bench mark .
in the end , come to pricing , GPT-4.1 cost $ 2/$8 for stimulant / output signal token per 1 million keepsake .
fundamentally , in medial inquiry , GPT-4.1 is 26 % punk than GPT-4o while offer much good functioning .
GPT-4.1 nano is the flash and fast exemplar by OpenAI , ever .
This was finally , openai cite that gpt-4.5 is lead aside from the api on fourteenth july , 2025 .
This was ## diving event into gpt-4.1
amount to benchmark , openai say on swe - bench verified , gpt-4.1 finish 55 % of labor , much high-pitched than gpt-4o ’s 33 % and o3 - mini - in high spirits ’s 49 % .
Next , in educational activity keep up , GPT-4.1 get along very near to GPT-4.5 , o3 - mini - gamey , and o1 - high-pitched on OpenAI ’s intragroup bench mark .
last , come to pricing , GPT-4.1 cost $ 2/$8 for stimulation / yield item per 1 million token .
fundamentally , in average interrogation , GPT-4.1 is 26 % chinchy than GPT-4o while offer up much good functioning .
GPT-4.1 nano is the crummy and fast mannequin by OpenAI , ever .
in conclusion , OpenAI mention that GPT-4.5 is break out from the API on fourteenth July , 2025 .
