Compare / Mistral Large 3 2512 vs GPT-5 Mini
Mistral Large 3 2512 vs GPT-5 Mini
21 results from 3 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | Mistral Large 3 2512 | GPT-5 Mini |
|---|---|---|---|
| Arena (formerly LMArena)reported | OverallArena rating | 1,434 | 1,394 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,409 | 1,381 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,373 | 1,324 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,428 | 1,400 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,400 | 1,373 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,413 | 1,390 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,388 | 1,352 |
| Artificial Analysisreported | Artificial Analysis Coding Indexindex score | 20.1 | 15.6 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 9.30 | 20.6 |
| Artificial Analysisreported | GPQA Diamond% of questions | 68.0% | 82.8% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 4.2% | 21.5% |
| Artificial Analysisreported | IFBench% of instructions | 36.2% | 75.4% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 36.0% | 72.3% |
| Artificial Analysisreported | SciCode% of problems | 36.6% | 39.0% |
| Artificial Analysisreported | Terminal-Bench 2.1% of tasks | 12.0% | 3.7% |
| Artificial Analysisreported | Terminal-Bench 4.0% of tasks | 0.0% | 0.0% |
| Artificial Analysisreported | Terminal-Bench Hard% of tasks | 15.9% | 33.3% |
| Artificial Analysisreported | Τ-bench banking% of tasks | 5.8% | 15.5% |
| Artificial Analysisreported | Τ²-bench telecom% of tasks | 24.6% | 71.1% |
| Vectara Hallucination Leaderboardreported | Answer rate% of documents | 98.8% | 99.9% |
| Vectara Hallucination Leaderboardreported | Hallucination rate% of summaries | 14.5% | 12.9% |
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.