Gemini 3 Flash (Preview) (Non-reasoning): benchmark results

Gemini 3 Flash (Preview) evaluated with reasoning disabled. Provider: Google. Released 2025-12-17. Access: API.

Aggregate Rating 1688 ± 11, rank #394 of 2170 rated models, from 329 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AgentProcessBench - HotPotQA Step Accuracy76.7StepAcc (%): share of assistant steps, pooled over trajector100
Doc-PP - Implicit Query Faithfulness (With Policy)69Faithfulness (%; content-checklist criteria met, implicit qu100
EuroEval Albanian NLU - MMS SQ32.13Matthews correlation coefficient (%)100
EuroEval Bulgarian Knowledge80.16Knowledge Average Score (%)100
EuroEval Czech Summarization - Czech News30.2chrF3++ (%)100
EuroEval Estonian NLU - Grammar ET66.32Matthews correlation coefficient (%)100
EuroEval Finnish Summarization - Xlsum FI32.86chrF3++ (%)100
nsanku-MMLU - Akuapem Twi66.5Accuracy (%)100
nsanku-MMLU - Asante Twi68Accuracy (%)100
nsanku-MMLU - Dagaare47Accuracy (%)100
nsanku-MMLU - Dagbani47Accuracy (%)100
nsanku-MMLU - Dangme47Accuracy (%)100

Interactive version: theaggregate.ai/model?slug=gemini-3-flash-preview-non-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-05.