An X post from @chetaslua reports that Opus 5.1 correctly answered the Opus 4.6 release-date question in a fresh chat: Thursday, February 5, 2026. The post says the model had no tools, memory, or search available, and contrasts that answer with responses from a base Opus 5 model that either failed to answer or drifted to November 2025 or January 2026.

Original post on X

Load the post to view it as published on X. X may receive connection data.

View the original post on X ↗

A precise release-date recall test

The exchange focused on a narrow question: when exactly did Opus 4.6 release? According to the post, Opus 5.1 returned both the date and the weekday. The post identifies February 5, 2026 as the shipping date and points to Anthropic’s SDK commit a9f163fa as supporting evidence. February 5, 2026 was a Thursday.

The detail matters because the question tests more than whether a model recognizes the Opus product line. It asks for a specific historical fact and an exact calendar date, leaving less room for a vague or approximate answer.

How the answer differed from base Opus 5

The post presents the result as a contrast with base Opus 5. It says that model either blanked on the question or supplied a date in November 2025 or January 2026. The reported difference is therefore specific: Opus 5.1 produced the expected release date, while the base model reportedly showed uncertainty or date drift.

That makes the exchange a useful anecdote about version-history recall. It does not establish broader accuracy across all subjects, but it does show the kind of small factual distinction that can affect confidence in an answer about when a model became available.

What the comparison indicates

For release-history questions, a confident-sounding response is not enough. The useful answer needs to identify the exact date and, where available, point to a concrete reference. In this case, the post’s cited SDK commit gives the reported February 5, 2026 date a specific supporting reference, while the comparison with base Opus 5 highlights how easily a model can otherwise drift to a nearby but incorrect timeframe.

The clearest takeaway is narrow: in this reported no-tools test, Opus 5.1 recalled the Opus 4.6 release date more accurately than the base Opus 5 responses described in the post.