Celeris-1: benchmark results

Provider: Celeris. Released 2026-07-23. Access: API.

Unified ELO 1489 ± 1, rank #921 of 1761 rated models, from 17 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA Humanity's Last Exam6.81Accuracy (%)44.3
BenchmarkList ECI106.56Capability Index (ECI)44.1
AA GPQA Diamond63.13Accuracy (%)38.3
Artificial Analysis Intelligence Index6.26Intelligence Index38.2
AA Long Context Reasoning38Accuracy (%)38
AA CritPt0Accuracy (%)24.4
AA GDPval485.2ELO19.8
AA Omniscience - Business10.51Accuracy (%)17
AA Omniscience - Science, Engineering & Mathematics16.7Accuracy (%)15.5
AA-Omniscience Accuracy10.98Accuracy (%)12.9
AA Omniscience - Software Engineering (SWE)13.33Accuracy (%)12.1
AA Omniscience - Health10.2Accuracy (%)11.9

Interactive version: theaggregate.ai/model?slug=celeris-1 · How It Works · Data refreshed daily, snapshot 2026-09-05.