maybe this faggot 5 will help me more with work

Is GPT‑5 better at engineering calculations?
Absolutely, it's a massive step forward. GPT‑5 significantly improves performance across logic-heavy and technical domains:

It sets new state-of-the-art results on rigorous math and reasoning benchmarks — including academic tests like the AIME (94.6% accuracy) and GPQA (88.4% with the “thinking” mode)
OpenAI
+1
.

It's more reliable, less prone to hallucinations, and better at following complex, multi-step instructions, which is crucial for engineering tasks
Wikipedia
+10
OpenAI
+10
The Verge
+10
.

Previously, studies on GPT‑4 and earlier models showed limitations in engineering calculations—ChatGPT sometimes faltered on nuanced or under-specified problems, especially in fields like statics
OpenAI Community
+4
arXiv
+4
arXiv
+4
. Those shortcomings are much less pronounced in GPT‑5, thanks to its enhanced reasoning capabilities and benchmark-leading performance.