gpt-6IMAGE
Grok 4.7 scores high but has no comparison with GPT-6 Astra
A Japanese post says Grok 4.7 at maximum reasoning effort outscores Opus 5 but falls short of Fable 5.1, and for unspecified reasons there is no comparison with GPT-6 Astra.
Kinopee
@kinopee_aiArchived 2026.09.22Originally posted 2026.09.22
Kinopee @kinopee_aiOriginally posted 2026.09.22
Grok 4.7、reasoning effort を一番高く設定した状態だと Opus 5 よりも高スコア。Fable 5.1 にはさすがに叶わないのと GPT-6 Astra との比較はない（大人の事情
x.com/kinopee_ai/status/2102071792370332021Archived 2026.09.22 10:18