Reaction
Questions about 1017's cost savings and benchmark performance
A user questions 1017's cost savings versus V4.1 Flash, noting similar verbosity in AA's figures.
TLDR
In two replies, a user says 1017 has a place on the speed Pareto frontier but recommends “Sol juice” instead. They question its cost savings versus V4.1 Flash, citing similar verbosity in AA's figures and asking whether AA used off-peak pricing. They call the benchmark results modest: they say it's the best DS model on most benchmarks, including CritPt, which they consider broken, but are skeptical of its apparent loss to MiMo.
Combined views
600
1 Source, first seen ago
