🏅 Top traders

My source is saying "obviously not" but also I'm the main NO holder and it feels bad so I will seek other opinions

Ok my second source is saying "No"

I'll resolve this No now but i strongly encourage disagreement if someone thinks this is actually worth investigating more / if you think an independent investigation wouldn't have come to the same conclusion

Tough to resolve because everyone went from opus 4.5 to opus 4.6, so there was very little comparison between opus 4.5 and sonnet 4.6. But this still seems right.

@comicstosteal yeah that was my initial realization and i thought maybe i'd need to N/A but it seems the powerusers in question at least tried sonnet 4.6 enough to know where they stood on this question

This market has been closed for ~42 hours. Could the creator resolve? The question period seems to have concluded.

new version for claude 5

I don't understand why this is settling at 40% odds but that has 70%

@comicstosteal conditional on their naming their next release 4.6, it will likely be decently good. Conditional on them naming it 5, it will likely be very good

oh darn it'll resolve N/A

Sonnet 4.5 was roughly on par or slightly better than Opus 4 and Opus 4.1.

So my guess would be that Sonnet 5 is roughly on par or slightly better than Opus 4.5. But not Sonnet 4.6

based and correcttake-pilled

Does This resolve N/A if there's no model called "Sonnet 4.6" release by January 31? And/or why do you think such a model will exist?

As a super user. My personal perspective was that Claude opus 4.1 was far better at coding than sonnet 4.5. So I’m curious how you’re going to determine this.

@StatsMan ah. hmmmmmmm

@Bayesian Do a poll?

@JoshSnider yeah sure but not of random manifold users, i'll poll full time coders that make use of claude max 5x or 20x and what their sense is

Just want to clarify that if the next release was .7 or .75 that would count

@BenAybar If they go straight to Sonnet 5?

Well I don't decide because it's not my market but I would imagine it should n/a

@BenAybar Oh whoops sorry I took your original post as a statement not a question.

@BenAybar just sonnet 4.6. If it is never released, resolves N/A. If it's released after january, resolution date is pushed back. does that work?

@Bayesian yeah that's fine

How are you gonna find out what power users who don't concern themselves with model cost think?

@AlexanderTheGreater i know a couple

and i'm hoping ppl yap enough on twitter that it's unambiguous