Hermes 4 70B (Non-reasoning) — benchmark results

Provider: Nous Research. Released 2025-08-26. Access: Open.

Unified ELO 1507 ± 39, rank #762 of 1756 rated models, from 36 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA Omniscience - Software Engineering (SWE) - Julia20Accuracy (%)65.4
AA Omniscience - Humanities & Social Sciences22.7Accuracy (%)61.8
AA Omniscience - Software Engineering (SWE) - PHP28Accuracy (%)59.7
AA Omniscience - Software Engineering (SWE) - Dart22Accuracy (%)54.8
AA Omniscience - Law11.3Accuracy (%)52.2
AA Omniscience - Software Engineering (SWE) - R12Accuracy (%)52.1
AA Omniscience - Health19.2Accuracy (%)51.5
AA Omniscience - Business16.3Accuracy (%)48.9
AA Omniscience - Software Engineering (SWE) - Python20Accuracy (%)48.7
AA Omniscience - Software Engineering (SWE) - Swift36Accuracy (%)48
AA-Omniscience Accuracy18.18Accuracy (%)45.7
AA Omniscience - Software Engineering (SWE) - C32Accuracy (%)44

Interactive version: aibenchmarks.dev/model?slug=hermes-4-70b-non-reasoning · How the rankings work · Data refreshed daily, snapshot 2026-07-17.