Compare / GPT-5 Mini vs GLM 5.1
GPT-5 Mini vs GLM 5.1
21 results from 4 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | GPT-5 Mini | GLM 5.1 |
|---|---|---|---|
| Arena (formerly LMArena)reported | OverallArena rating | 1,394 | 1,456 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,381 | 1,465 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,324 | 1,448 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,400 | 1,495 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,373 | 1,460 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,390 | 1,465 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,352 | 1,453 |
| Artificial Analysisreported | Artificial Analysis Coding Indexindex score | 15.6 | 55.8 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 16.8 | 26.1 |
| Artificial Analysisreported | GPQA Diamond% of questions | 82.8% | 86.8% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 21.5% | 30.1% |
| Artificial Analysisreported | IFBench% of instructions | 75.4% | 76.3% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 72.3% | 73.7% |
| Artificial Analysisreported | SciCode% of problems | 39.0% | 44.8% |
| Artificial Analysisreported | Terminal-Bench 2.1% of tasks | 3.7% | 61.8% |
| Artificial Analysisreported | Terminal-Bench 4.0% of tasks | 0.0% | 2.0% |
| Artificial Analysisreported | Terminal-Bench Hard% of tasks | 33.3% | 43.2% |
| Artificial Analysisreported | Τ-bench banking% of tasks | 15.5% | 13.6% |
| Artificial Analysisreported | Τ²-bench telecom% of tasks | 71.1% | 97.7% |
| SimpleQA Verified (Epoch AI)reported | Correct answers% of questions | 21.6% | 34.0% |
| Vending-Bench 2reported | Money after a yearUS dollars | $-31.18 | $5,634.41 |
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.