Compare / DeepSeek V3.2 vs Grok 4.3
DeepSeek V3.2 vs Grok 4.3
19 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 V3.2 | Grok 4.3 |
|---|---|---|---|
| Arena (formerly LMArena)reported | OverallArena rating | 1,437 | 1,428 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,423 | 1,440 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,400 | 1,423 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,453 | 1,452 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,420 | 1,415 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,425 | 1,442 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,410 | 1,424 |
| Artificial Analysisreported | Artificial Analysis Coding Indexindex score | 44.2 | 42.2 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 21.5 | 24.9 |
| Artificial Analysisreported | GPQA Diamond% of questions | 84.0% | 90.1% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 24.6% | 37.2% |
| Artificial Analysisreported | IFBench% of instructions | 60.7% | 83.3% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 73.3% | 75.0% |
| Artificial Analysisreported | Terminal-Bench 2.1% of tasks | 46.8% | 39.7% |
| Artificial Analysisreported | Terminal-Bench Hard% of tasks | 35.6% | 37.9% |
| Artificial Analysisreported | Τ²-bench telecom% of tasks | 90.6% | 97.7% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 26.0% | 21.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.36 | 0.34 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 57.3 | 57.7 |
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.