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 1688 ± 33, rank #222 of 1640 rated models, from 69 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA Long Context Reasoning | 83.33 | Accuracy (%) | 96.3 |
| AA-Omniscience Hallucination Rate | 35.83 | Hallucination Rate (%) | 88.8 |
| Artificial Analysis Intelligence Index | 24.05 | Intelligence Index | 73.9 |
| AA-Omniscience Index - Law | -7.5 | Omniscience Index | 73.4 |
| AA Humanity's Last Exam | 25.76 | Accuracy (%) | 73 |
| AA-Omniscience Index - Business | -8 | Omniscience Index | 72.3 |
| AA-Omniscience Index - Humanities & Social Sciences | -7.6 | Omniscience Index | 70.1 |
| AA-Omniscience Index - Science, Engineering & Mathematics | -6.1 | Omniscience Index | 68.9 |
| LLM Stats Score | 31.94 | LLM Stats Score (conservative rating) | 68.1 |
| AA Omniscience | -10.82 | Score | 65.6 |
| AA-Omniscience Index - Health | -20.1 | Omniscience Index | 63.7 |
| AA CritPt | 1.43 | Accuracy (%) | 62.7 |
Interactive version: theaggregate.ai/model?slug=solar-mini-4 · How It Works · Data refreshed daily, snapshot 2026-10-13.