Compare / Ministral 3 8B 2512 vs Kimi K2.6
Ministral 3 8B 2512 vs Kimi K2.6
17 results from 3 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | Ministral 3 8B 2512 | Kimi K2.6 |
|---|---|---|---|
| Artificial Analysisreported | Artificial Analysis Coding Indexindex score | 9.70 | 61.8 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 5.50 | 27.0 |
| Artificial Analysisreported | GPQA Diamond% of questions | 47.1% | 91.1% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 4.3% | 37.5% |
| Artificial Analysisreported | IFBench% of instructions | 29.1% | 76.0% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 25.7% | 81.0% |
| Artificial Analysisreported | SciCode% of problems | 20.7% | 51.5% |
| Artificial Analysisreported | Terminal-Bench 2.1% of tasks | 4.1% | 65.9% |
| Artificial Analysisreported | Terminal-Bench 4.0% of tasks | 0.0% | 0.5% |
| Artificial Analysisreported | Terminal-Bench Hard% of tasks | 4.5% | 43.9% |
| Artificial Analysisreported | Τ-bench banking% of tasks | 3.7% | 23.3% |
| Artificial Analysisreported | Τ²-bench telecom% of tasks | 26.6% | 95.9% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 23.0% | 18.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.33 | 0.35 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 28.8 | 64.4 |
| Vectara Hallucination Leaderboardreported | Answer rate% of documents | 99.1% | 99.7% |
| Vectara Hallucination Leaderboardreported | Hallucination rate% of summaries | 21.7% | 10.8% |
Bold green marks the better value on that metric. Where a source reports ranges that overlap, the difference may not be meaningful; see the benchmark page for ranges.