modelbenchmark.io

llama-4-scout-17b-16e-instruct-fp8

Meta · llama-4-scout-17b-16e-instruct-fp8

Compare

Specification

most-agreed values

Context
10M
Max output
4K
Released
—
Knowledge cutoff
—
Retires
—
Open weights
—
Input
—
Output
—

Available from 1 host

HostIn $/MOut $/MCache rdCache wrContextOutputRetires
meta_llama · meta-llama————10M4K—

More from Meta

most-hosted first