Gemma 4 31B Instruct

Google llm

Google's flagship dense Gemma 4 model, natively multimodal with a 256K context, released under Apache 2.0 instead of the prior custom Gemma licence.

overall ⌄

57.9

26 of 93 ranked

price · $/M tokens

self-hosted

open weights, no first-party API

to run it yourself ⌄

19 GB

minimum · 74 GB recommended

Benchmarks

What its maker published, and what anyone else measured. Every figure links the document it came from.

benchmark	published	measured
MMLU-Pro	85.2	—
GPQA	84.3	—

No recorded lineage — root or standalone model.