GPT-6 Astra (Low): benchmark results
Provider: OpenAI. Released 2026-09-03. Access: API.
Unified ELO 1938 ± 11, rank #28 of 2072 rated models, from 109 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA-Omniscience Index - Software Engineering (SWE) - JavaScript | 87.27 | Omniscience Index | 99.4 |
| AA-Omniscience Index - Business | 33 | Omniscience Index | 98.8 |
| AA-Omniscience Index - Software Engineering (SWE) - Go | 88 | Omniscience Index | 98.6 |
| AA-Omniscience Index - Health | 22.7 | Omniscience Index | 98.4 |
| ParseBench - Tables | 93.17 | GTRM Score (GriTS + TableRecordMatch) | 98.1 |
| AA-Omniscience Index - Software Engineering (SWE) - Rust | 84 | Omniscience Index | 98 |
| AA-Omniscience Index - Software Engineering (SWE) - TypeScript | 91.11 | Omniscience Index | 98 |
| AA-Omniscience Index - Humanities & Social Sciences | 38.9 | Omniscience Index | 97.9 |
| AA-Omniscience Index - Software Engineering (SWE) | 80.9 | Omniscience Index | 97.5 |
| AA-Omniscience Index - Software Engineering (SWE) - C | 88 | Omniscience Index | 97.4 |
| AA-Omniscience Index - Software Engineering (SWE) - Java | 67 | Omniscience Index | 97.4 |
| AA Omniscience | 40.55 | Score | 97.3 |
Interactive version: theaggregate.ai/model?slug=gpt-6-astra-low · How It Works · Data refreshed daily, snapshot 2026-10-02.