Compare / DeepSeek V3.2 Exp vs Ministral 3 8B 2512
DeepSeek V3.2 Exp vs Ministral 3 8B 2512
12 results from 3 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | DeepSeek V3.2 Exp | Ministral 3 8B 2512 |
|---|---|---|---|
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 16.6 | 5.50 |
| Artificial Analysisreported | GPQA Diamond% of questions | 79.7% | 47.1% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 14.9% | 4.3% |
| Artificial Analysisreported | IFBench% of instructions | 54.1% | 29.1% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 72.3% | 25.7% |
| Artificial Analysisreported | Terminal-Bench Hard% of tasks | 31.1% | 4.5% |
| Artificial Analysisreported | Τ²-bench telecom% of tasks | 33.9% | 26.6% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 25.0% | 23.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.35 | 0.33 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 54.0 | 28.8 |
| Vectara Hallucination Leaderboardreported | Answer rate% of documents | 96.6% | 99.1% |
| Vectara Hallucination Leaderboardreported | Hallucination rate% of summaries | 5.3% | 21.7% |
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.