Compare / Gemini 3.1 Pro Preview vs Grok 4.5
Gemini 3.1 Pro Preview vs Grok 4.5
31 results from 8 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | Gemini 3.1 Pro Preview | Grok 4.5 |
|---|---|---|---|
| APEX-Agentsreported | Tasks passed% of tasks | 35.3% | 56.2% |
| Arena (formerly LMArena)reported | Confirmed task successIPS effect estimate | -0.08 | -0.01 |
| Arena (formerly LMArena)reported | Praise over complaintIPS effect estimate | -0.01 | -0.03 |
| Arena (formerly LMArena)reported | SteerabilityIPS effect estimate | -0.04 | 0.04 |
| Arena (formerly LMArena)reported | Tool groundingIPS effect estimate | -0.00 | 0.00 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,459 | 1,464 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,208 | 1,202 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,472 | 1,465 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,477 | 1,465 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,480 | 1,452 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,509 | 1,488 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,479 | 1,462 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,487 | 1,465 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,480 | 1,459 |
| Artificial Analysisreported | Artificial Analysis Coding Indexindex score | 68.8 | 72.4 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 29.7 | 38.8 |
| Artificial Analysisreported | GPQA Diamond% of questions | 94.1% | 93.1% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 47.0% | 42.7% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 82.0% | 79.3% |
| Artificial Analysisreported | SciCode% of problems | 58.7% | 55.0% |
| Artificial Analysisreported | Terminal-Bench 2.1% of tasks | 73.8% | 81.6% |
| Artificial Analysisreported | Terminal-Bench 4.0% of tasks | 4.0% | 10.6% |
| Artificial Analysisreported | Τ-bench banking% of tasks | 21.4% | 42.1% |
| GDP.pdfreported | Professional document tasks% of rubric | 17.0% | 14.0% |
| SimpleQA Verified (Epoch AI)reported | Correct answers% of questions | 73.5% | 48.3% |
| tau2-benchreported | Consistency (pass^4)% of tasks | 9.3% | 32.0% |
| tau2-benchreported | Task success (pass^1)% of tasks | 26.0% | 47.9% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 19.0% | 21.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.34 | 0.32 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 72.2 | 60.0 |
| Vending-Bench 2reported | Money after a yearUS dollars | $911.21 | $3,887.43 |
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.