Opus 5.5 is weirdly human to interact with
A user describes their experience with Opus 5.5, noting its impressive ability to maintain context over a long chat with numerous compactions. They suggest its performance might even surpass fable 5.1. The user expresses surprise, stating that Opus 5.5 is "weirdly human to interact with" and calls it the "biggest surprise in like a year or more of using an LLM."
This user account is one of the first to directly compare Opus 5.5's performance to fable 5.1, noting its superior context retention over long interactions.
Time & source
Times shown in UTC
Display time zone: UTC
Local time zone unavailable; showing UTC.
PublishedOffset at this time: UTC+0Sep 28, 2026, 22:45 UTC
IngestedOffset at this time: UTC+0Sep 29, 2026, 02:00 UTC
- Published
- Sep 28, 2026, 22:45
- Ingested
- Sep 29, 2026, 02:00
- Source type
- Dev community
- Tier
- Community
- Source status
- Healthy
Tier is a per-source editorial setting, not a per-item score.
Discussion trend
The percentage is based on collected discussion signal, not new comments or independent people. The curve only compares the same topic across time.
I have a quite long running chat with quite numerous compactions. It doesn't lose context, performs really well, maybe better than fable 5.1 (wouldn't know as I couldn't go so much with it).
But the part that strikes me the most is that interactions are almost human. Like it is working on a task, and in the middle I will try to interrupt with some questions follow-ups. What happens is that it will give a short answer, continue on the important details until a certain part of workflow is finished then bam remember my earlier question and goes on answering it fuller. Instead of fully derailing the work, it just seems to flow naturally with a really decent outcome just nudging the direction bit here bit there. It doesn't need handholding with a pause every step waiting for me to give directions, rather it can just continue going and expects me to go along with it on the side.
Usually I would rant about limits etc, but this is probably the biggest surprise in like a year or more of using an LLM