Compare / DeepSeek V4 Pro 0423 vs gemini-3-flash
DeepSeek V4 Pro 0423 vs gemini-3-flash
18 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 V4 Pro 0423 | gemini-3-flash |
|---|---|---|---|
| Arena (formerly LMArena)reported | OverallArena rating | 1,455 | 1,468 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,460 | 1,469 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,446 | 1,457 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,480 | 1,496 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,452 | 1,458 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,458 | 1,473 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,451 | 1,461 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 36.0 | 26.3 |
| Artificial Analysisreported | GPQA Diamond% of questions | 92.8% | 89.8% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 41.0% | 36.6% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 80.3% | 78.0% |
| Artificial Analysisreported | Τ-bench banking% of tasks | 39.6% | 20.8% |
| OpenHands Indexreported | Average% resolved | 40.7% | 49.0% |
| OpenHands Indexreported | Front end (SWE-Bench Multimodal)% resolved | 36.8% | 22.1% |
| OpenHands Indexreported | Greenfield (Commit0)% resolved | 12.5% | 18.8% |
| OpenHands Indexreported | Information gathering (GAIA)% resolved | 12.7% | 58.8% |
| OpenHands Indexreported | Issue resolution (SWE-Bench)% resolved | 73.2% | 74.6% |
| OpenHands Indexreported | Testing (SWT-Bench)% resolved | 68.1% | 70.7% |
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.