According to a recent Twitter post, Kimi K3 is apparently better than Opus 4.8 for some tasks: wh (@nrehiew_) on X

wh (@nrehiew_) on X

Here are a few benchmark scores of K3 that have been officially confirmed This is a Fable/Sol class model that is strictly better than Opus 4.8 across the board at Sonnet pricing. Insane

This market resolves to Yes if Zvi (in his blog or Twitter or a comment on this market) directly states that Kimi K3 is indeed better than Opus 4.8 for at least some (non trivial/edge case) tasks, such as coding.

Jokes/humor don't count, the market will resolve based off the spirit of the question.

Market resolves by the end of July 2026, can be extended if and only if Zvi doesn't do a write up about Kimi K3 before then.