Compare / Gemini 3.8 Flash vs GPT-5.4 Mini
Gemini 3.8 Flash vs GPT-5.4 Mini
27 results from 4 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | Gemini 3.8 Flash | GPT-5.4 Mini |
|---|---|---|---|
| BulletBenchmeasured by us | Cost per gameUS dollars | $0.0154 | $0.0205 |
| BulletBenchmeasured by us | Games lost on time% of games | 50.0% | 50.0% |
| BulletBenchmeasured by us | Invalid moves% of moves | 0.0% | 0.0% |
| BulletBenchmeasured by us | Ladder Eloladder Elo | 1,068 | 197 |
| BulletBenchmeasured by us | Median move timemilliseconds | 1.8 s | 0.9 s |
| BulletBenchmeasured by us | Cost per gameUS dollars | $0.0041 | $0.0136 |
| BulletBenchmeasured by us | Games lost on time% of games | 83.3% | 0.0% |
| BulletBenchmeasured by us | Invalid moves% of moves | 0.0% | 0.0% |
| BulletBenchmeasured by us | Ladder Eloladder Elo | 197 | 571 |
| BulletBenchmeasured by us | Median move timemilliseconds | 1.8 s | 0.9 s |
| Arena (formerly LMArena)reported | OverallArena rating | 1,477 | 1,447 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,482 | 1,460 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,482 | 1,402 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,529 | 1,480 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,488 | 1,434 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,492 | 1,447 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,486 | 1,424 |
| Artificial Analysisreported | Artificial Analysis Coding Indexindex score | 76.3 | 56.1 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 40.9 | 24.1 |
| Artificial Analysisreported | GPQA Diamond% of questions | 95.3% | 87.5% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 47.8% | 28.1% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 84.0% | 77.0% |
| Artificial Analysisreported | SciCode% of problems | 56.6% | 52.1% |
| Artificial Analysisreported | Terminal-Bench 2.1% of tasks | 87.6% | 59.2% |
| Artificial Analysisreported | Terminal-Bench 4.0% of tasks | 19.7% | 2.0% |
| Artificial Analysisreported | Τ-bench banking% of tasks | 45.8% | 25.6% |
| SimpleQA Verified (Epoch AI)reported | Correct answers% of questions | 69.7% | 29.4% |
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.