Celeris-1: benchmark results
Provider: Celeris. Released 2026-07-23. Access: API.
Aggregate Rating 1474 ± 20, rank #1172 of 2168 rated models, from 64 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA MMLU-Pro | 77.99 | Accuracy (%) | 58.9 |
| AA Humanity's Last Exam | 6.81 | Accuracy (%) | 40 |
| AA GPQA Diamond | 63.13 | Accuracy (%) | 38.1 |
| Arcee MMLU-Pro (BenchLM) | 75.9 | Accuracy (%) | 36.3 |
| AA Long Context Reasoning | 38 | Accuracy (%) | 33.9 |
| AA-Omniscience Index - Software Engineering (SWE) - Go | -64 | Omniscience Index | 22.9 |
| AA GDP.pdf - Engineering - Criterion Pass Rate | 28.69 | Criterion Pass Rate (%) | 22.7 |
| AA CritPt | 0 | Accuracy (%) | 22 |
| AA GDP.pdf - Insurance - Criterion Pass Rate | 50.78 | Criterion Pass Rate (%) | 21.3 |
| AutomationBench-AA - Strict | 0.15 | Strict Task Success (%) | 18.6 |
| AA Terminal-Bench 2.1 | 11.24 | Pass@1 (%) | 17.8 |
| AA GDP.pdf - Real Estate - Criterion Pass Rate | 46.04 | Criterion Pass Rate (%) | 17.3 |
Interactive version: theaggregate.ai/model?slug=celeris-1 · How It Works · Data refreshed daily, snapshot 2026-10-20.