gemma-4-31b-it Benchmark & Insights

Google Gemini API

Updated Jul 18, 2026 All models

Sample size

157 runs

in window

Accuracy

88.8%

consensus match · 160d

Confidence

100%

over 162 runs

Window end

Jul 18, 2026

most recent run

Input price

$0.13/MTok

prompt tokens

Output price

$0.38/MTok

completion tokens

Model insights

01 An outlier among open models: decent risk accuracy but the worst "safe" score in the top half, with ten false "unsafe" alarms on calm days.
02 It is firmly "pessimistic" (20 over vs 2 under), yet at $0.13/$0.38 it is one of the cheapest ways to get near-90% agreement.

Notes

Open-weight model; also available free via OpenRouter

Recent forecasts

Date

Conf.

Risk

Safe