Gemma 4 26B — benchmark results

Google Gemma 4 26B model row. Provider: Google. Released 2026-04-02. Access: API.

Unified ELO 1641 ± 50, rank #390 of 1776 rated models, from 18 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AgentFloor59.9TCR (self-reported)96.9
SEA-HELM70.82Mean Score (%)93
Wordle Arena83.33Win Rate (%)85.4
Fibble5 Arena9.09Win Rate (%)81.1
TeleResilienceBench27.2Macro Average CFR (self-reported)71.4
DiagnosticIQ63.2Macro % D.IQ (self-reported)65.6
Fibble Arena8.33Win Rate (%)64.1
SLR-Bench - Medium23Accuracy (%)57.3
Polaris-Bench47.2Overall C (self-reported)53.8
Wolfram LLM Benchmarking Project41.8Correct Functionality (%)52.7
SLR-Bench - Hard4Accuracy (%)51.8
AgentCIBench67.5Leakage (self-reported)50

Interactive version: theaggregate.ai/model?slug=gemma-4-26b · How the rankings work · Data refreshed daily, snapshot 2026-07-22.