Inkling Small (Max): benchmark results

Provider: Thinking Machines. Released 2026-07-30. Access: Open.

Aggregate Rating 1735 ± 24, rank #294 of 2245 rated models, from 72 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Nejumi 4 - Toxicity - Prohibited Acts97.92Criteria met (%)94.2
Nejumi 4 - Toxicity - Violation Categories54.76Criteria met (%)91.9
Nejumi 4 - MT-Bench (Japanese) - Humanities100Judge rating (1-10, x10)90.3
Nejumi 4 - MT-Bench (Japanese) - STEM100Judge rating (1-10, x10)89.6
Nejumi 4 - MT-Bench (Japanese) - Math100Judge rating (1-10, x10)89.3
Nejumi 4 - jaster (0-shot) - JSQuAD81.31Character F1 (x100)89
Nejumi 4 - MT-Bench (Japanese) - Extraction97.5Judge rating (1-10, x10)88.6
Nejumi 4 - jaster (0-shot) - MMLU-ProX (Japanese)90Exact match (%)88.3
Nejumi 4 - M-IFEval (Japanese)89.38Score (%)85.7
Nejumi 4 - jaster (0-shot) - JBLiMP96Exact match (%)84.4
Nejumi 4 - ARC-AGI-196Accuracy (%)81.2
Nejumi 4 - jaster (2-shot) - JSQuAD80.84Character F1 (x100)81.2

Interactive version: theaggregate.ai/model?slug=inkling-small-max · How It Works · Data refreshed daily, snapshot 2026-11-16.