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
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Nejumi 4 - Toxicity - Prohibited Acts | 97.92 | Criteria met (%) | 94.2 |
| Nejumi 4 - Toxicity - Violation Categories | 54.76 | Criteria met (%) | 91.9 |
| Nejumi 4 - MT-Bench (Japanese) - Humanities | 100 | Judge rating (1-10, x10) | 90.3 |
| Nejumi 4 - MT-Bench (Japanese) - STEM | 100 | Judge rating (1-10, x10) | 89.6 |
| Nejumi 4 - MT-Bench (Japanese) - Math | 100 | Judge rating (1-10, x10) | 89.3 |
| Nejumi 4 - jaster (0-shot) - JSQuAD | 81.31 | Character F1 (x100) | 89 |
| Nejumi 4 - MT-Bench (Japanese) - Extraction | 97.5 | Judge rating (1-10, x10) | 88.6 |
| Nejumi 4 - jaster (0-shot) - MMLU-ProX (Japanese) | 90 | Exact match (%) | 88.3 |
| Nejumi 4 - M-IFEval (Japanese) | 89.38 | Score (%) | 85.7 |
| Nejumi 4 - jaster (0-shot) - JBLiMP | 96 | Exact match (%) | 84.4 |
| Nejumi 4 - ARC-AGI-1 | 96 | Accuracy (%) | 81.2 |
| Nejumi 4 - jaster (2-shot) - JSQuAD | 80.84 | Character F1 (x100) | 81.2 |
Interactive version: theaggregate.ai/model?slug=inkling-small-max · How It Works · Data refreshed daily, snapshot 2026-11-16.