Mi:dm K 2.5 Pro Preview: benchmark results
Preview of KT's Mi:dm K 2.5 Pro, a 32B reasoning-focused Korean enterprise flagship strong in agentic tool use. Provider: KT. Released 2025-12-11. Access: API.
Aggregate Rating 1574 ± 41, rank #1252 of 10870 rated models, from 14 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA AIME 2025 | 78.67 | Accuracy (%) | 73.8 |
| AA MMLU-Pro | 81.27 | Accuracy (%) | 73.7 |
| AA LiveCodeBench | 57.57 | Pass@1 (%) | 63.6 |
| Tau2-Bench Telecom | 49.42 | Success Rate (%) | 53.9 |
| AA IFBench | 45.58 | Accuracy (%) | 52.4 |
| AA TAU-2 Bench | 49.42 | Accuracy (%) | 52.4 |
| AA GPQA Diamond | 72.22 | Accuracy (%) | 51.6 |
| AA Humanity's Last Exam | 9.08 | Accuracy (%) | 45.9 |
| AA-Omniscience Accuracy | 16.42 | Accuracy (%) | 28.8 |
| AA Terminal-Bench Hard | 3.03 | Accuracy (%) | 22.9 |
| AA CritPt | 0 | Accuracy (%) | 20.7 |
| AA Long Context Reasoning | 14.67 | Accuracy (%) | 17 |
Interactive version: theaggregate.ai/model?slug=mi-dm-k-2-5-pro-preview · How It Works · Data refreshed daily, snapshot 2026-11-12.