JT-4.1 Flash 236B A21B: benchmark results
Provider: China Mobile. Released 2026-07-09. Access: API.
Aggregate Rating 1762 ± 34, rank #153 of 10916 rated models, from 36 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA Humanity's Last Exam | 43.56 | Accuracy (%) | 91.2 |
| AA CritPt | 14 | Accuracy (%) | 81.3 |
| AA Long Context Reasoning | 78.33 | Accuracy (%) | 79 |
| AA GPQA Diamond | 84.55 | Accuracy (%) | 76.7 |
| AA GDPval | 1335.19 | ELO | 66.6 |
| AutomationBench-AA - Support - Raw Objective Completion | 82.23 | Raw Objective Completion (%) | 62.8 |
| AutomationBench-AA - Strict | 17.81 | Strict Task Success (%) | 61.5 |
| AA SciCode | 51.97 | Accuracy (%) | 61.2 |
| AutomationBench-AA - Operations - Raw Objective Completion | 80.97 | Raw Objective Completion (%) | 57.9 |
| AA Terminal-Bench 2.1 | 59.55 | Pass@1 (%) | 57.4 |
| AutomationBench-AA | 46.47 | Guardrail-Adjusted Objective Completion (%) | 56.7 |
| AutomationBench-AA - Finance - Raw Objective Completion | 71.98 | Raw Objective Completion (%) | 54.7 |
Interactive version: theaggregate.ai/model?slug=jt-4-1-flash-236b-a21b · How It Works · Data refreshed daily, snapshot 2026-11-08.