Motif-2-12.7B (Reasoning): benchmark results

Motif-2-12.7B evaluated with reasoning enabled. Provider: Motif. Released 2025-11-07. Access: Open.

Unified ELO 1503 ± 1, rank #837 of 1761 rated models, from 17 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA IFBench57.01Accuracy (%)68.3
AA Humanity's Last Exam8.76Accuracy (%)50.8
AA TAU-2 Bench46.49Accuracy (%)50.2
AA GPQA Diamond69.49Accuracy (%)47.4
Artificial Analysis Intelligence Index6.96Intelligence Index40.8
BenchmarkList ECI102.77Capability Index (ECI)39.4
AA Omniscience - Law9.62Accuracy (%)37.4
AA Omniscience - Humanities & Social Sciences16.68Accuracy (%)35.3
AA Omniscience - Science, Engineering & Mathematics23.2Accuracy (%)31.7
AA Omniscience - Health15.5Accuracy (%)29.3
AA-Omniscience Accuracy15.17Accuracy (%)25.8
AA Terminal-Bench Hard3.79Accuracy (%)25.6

Interactive version: theaggregate.ai/model?slug=motif-2-12-7b-reasoning · How It Works · Data refreshed daily, snapshot 2026-09-05.