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 1568 ± 32, rank #572 of 1776 rated models, from 36 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA AIME 202580.33Accuracy (%)75.9
AA LiveCodeBench65.08Pass@1 (%)71.8
AA IFBench57.01Accuracy (%)68.2
AA MMLU-Pro79.63Accuracy (%)65.1
AA Humanity's Last Exam8.2Accuracy (%)55.5
AA GPQA Diamond69.49Accuracy (%)51.7
AA TAU-2 Bench46.49Accuracy (%)50.1
AA Omniscience - Software Engineering (SWE) - Dart20Accuracy (%)47.9
Artificial Analysis Intelligence Index12.79Intelligence Index43.9
AA Omniscience - Law9.8Accuracy (%)42.4
AA Omniscience - Software Engineering (SWE) - HTML28Accuracy (%)41.5
AA Omniscience - Software Engineering (SWE) - Java15Accuracy (%)41.3

Interactive version: theaggregate.ai/model?slug=motif-2-12-7b-reasoning · How the rankings work · Data refreshed daily, snapshot 2026-07-22.