Compare / Gemini 3.5 Flash Lite vs GPT-5.4 Mini
Gemini 3.5 Flash Lite vs GPT-5.4 Mini
26 results from 3 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | Gemini 3.5 Flash Lite | GPT-5.4 Mini |
|---|---|---|---|
| BulletBenchmeasured by us | Cost per gameUS dollars | $0.0058 | $0.0205 |
| BulletBenchmeasured by us | Games lost on time% of games | 0.0% | 50.0% |
| BulletBenchmeasured by us | Invalid moves% of moves | 0.0% | 0.0% |
| BulletBenchmeasured by us | Ladder Eloladder Elo | 780 | 197 |
| BulletBenchmeasured by us | Median move timemilliseconds | 0.9 s | 0.9 s |
| BulletBenchmeasured by us | Cost per gameUS dollars | $0.0070 | $0.0136 |
| BulletBenchmeasured by us | Games lost on time% of games | 0.0% | 0.0% |
| BulletBenchmeasured by us | Invalid moves% of moves | 0.0% | 0.0% |
| BulletBenchmeasured by us | Ladder Eloladder Elo | 892 | 571 |
| BulletBenchmeasured by us | Median move timemilliseconds | 0.9 s | 0.9 s |
| Arena (formerly LMArena)reported | OverallArena rating | 1,448 | 1,447 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,454 | 1,460 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,435 | 1,402 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,467 | 1,480 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,444 | 1,434 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,456 | 1,447 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,442 | 1,424 |
| Artificial Analysisreported | Artificial Analysis Coding Indexindex score | 49.3 | 56.1 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 22.2 | 24.1 |
| Artificial Analysisreported | GPQA Diamond% of questions | 83.8% | 87.5% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 18.8% | 28.1% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 76.0% | 77.0% |
| Artificial Analysisreported | SciCode% of problems | 41.3% | 52.1% |
| Artificial Analysisreported | Terminal-Bench 2.1% of tasks | 53.6% | 59.2% |
| Artificial Analysisreported | Terminal-Bench 4.0% of tasks | 1.0% | 2.0% |
| Artificial Analysisreported | Τ-bench banking% of tasks | 17.5% | 25.6% |
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.