modelbenchmark.io

Google Gemma 4 31B Instruct

google-gemma-4-31b-it

Compare

Largest Gemma 4 instruction model for open, self-hosted chat and reasoning

Specification

most-agreed values

Context
256K
Max output
8K
Released
2026-04-03
Knowledge cutoff
Retires
Open weights
yes
Input
text, image, video
Output
text

Price

US dollars per million tokens · most-agreed

Input
$0.12
Output
$0.36
Cache read
$0.09

Available from 1 host

HostIn $/MOut $/MCache rdCache wrContextOutputRetires
venice$0.12$0.36$0.09256K8K