Gemini 3.8 Flash (Low): benchmark results

Provider: Google. Released 2026-09-02. Access: API.

Unified ELO 1704 ± 1, rank #91 of 1761 rated models, from 19 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA MMMU-Pro84.51Accuracy (%)95.7
AA Omniscience - Law57Accuracy (%)95.7
AA Omniscience - Humanities & Social Sciences52.7Accuracy (%)94
AA GPQA Diamond92.02Accuracy (%)93.3
AA-Omniscience Accuracy52.22Accuracy (%)93.2
AA Omniscience - Business43.8Accuracy (%)93.1
AA Omniscience21.35Score93
Artificial Analysis Intelligence Index42.37Intelligence Index92.2
AA Long Context Reasoning80.67Accuracy (%)91.8
AA Omniscience - Science, Engineering & Mathematics49Accuracy (%)91
AA Omniscience - Health42.8Accuracy (%)90.3
AA Omniscience - Software Engineering (SWE)68Accuracy (%)89.7

Interactive version: theaggregate.ai/model?slug=gemini-3-8-flash-low · How It Works · Data refreshed daily, snapshot 2026-09-05.