Solar Mini 4: benchmark results

Upstage's compact agent-oriented MoE (35B total, 3B active) with a 512K context, offered through the Upstage Console API as the successor to Solar Mini. Provider: Upstage. Released 2026-09-22. Access: API.

Aggregate Rating 1676 ± 19, rank #445 of 10556 rated models, from 75 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA Long Context Reasoning83.33Accuracy (%)95.9
AA-Omniscience Hallucination Rate35.83Hallucination Rate (%)88.4
AA Humanity's Last Exam25.76Accuracy (%)70.6
LLM Stats Score32.28LLM Stats Score (conservative rating)66.5
AA-Omniscience v1.1 Hallucination Rate36.53Hallucination Rate (%)65.7
AA Omniscience-10.82Score63
AA CritPt1.43Accuracy (%)58.6
AIIQ Composite IQ107Composite IQ (imported)48.6
AA GDPval1092.1ELO47.5
AA-Briefcase - PDF - Rubric Pass Rate31.03Rubric Pass Rate (%)44.4
AA SciCode47.57Accuracy (%)40.7
AA-Omniscience Index - Law-6.44Omniscience Index40.6

Interactive version: theaggregate.ai/model?slug=solar-mini-4 · How It Works · Data refreshed daily, snapshot 2026-11-17.