Is Claude Opus 4.5 Worth the Hype?
People are saying Claude Opus 4.5 is the next ChatGPT moment.
I did not know if I could trust the hype.
At Tenex, we help companies go from AI-absent to AI-native.
So I asked the team for honest reviews.
Our engineers at Tenex are maniacal about testing new models and optimizing their AI workflows, so I knew they would be real.
Here is what they said.
Table of Contents
Speed and Context
The first thing that stood out was the performance.
“Phenomenal so far. Faster than I expected for Opus and handles context really well.”
No complaints so far, and the consensus is that it only increases speed and quality.
One engineer told me they have already noticed themselves feeling more confident asking it to do tasks without providing step-by-step instructions because it seems more intentional with its planning.
Understanding Intent
The ability to understand what you actually want is where this model shines.
“Yeah, very, very good. It is clearly very smart but the thing I have loved about Anthropic across the board is how their models understand intent and explore.”
Opus 4.5 is very good at understanding intent and then being thorough.
If you say “do x” Sonnet 4.5 usually does x okay, but Opus 4.5 will look around a while until it knows how to do x your way.
It will even say something like, “Let me look for more examples.”
Complex Tasks
For the more challenging work, the difference becomes even more apparent.
“Yeah, especially for complex multi-turn tasks, and even for planning, noticing fewer hallucinations.”
Sometimes Sonnet 4.5 avoided guidelines like always leveraging existing libraries or code, and in some cases, it used to over-engineer.
But Opus 4.5’s instruction following also feels superior.
The Trade-offs
There is one clear downside worth mentioning.
“It is very smart but slow.”
One engineer has not used it to code yet, but said it can make a nice architecture diagram in a sales proposal.
Jokes aside, it is the first model that can consistently get the lines in an ASCII architecture diagram aligned correctly.
Conclusion
Based on my team’s feedback, Claude Opus 4.5 lives up to the hype for complex tasks that require understanding intent and thorough execution.
The trade-off is speed, but the quality improvement seems worth it for the right use cases.