Compare / Gemini 3.1 Pro Preview vs Kimi K3
Gemini 3.1 Pro Preview vs Kimi K3
35 results from 5 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 | Kimi K3 |
|---|---|---|---|
| BulletBenchmeasured by us | Ladder Eloladder Elo | 683 | 0 |
| BulletBenchmeasured by us | Median move timemilliseconds | 6.2 s | 6.8 s |
| CatalogBenchmeasured by us | Content quality% of checks | 96.6% | 95.8% |
| CatalogBenchmeasured by us | Failed outputs% of products | 0.0% | 0.6% |
| CatalogBenchmeasured by us | Publish-ready listings% of products | 3.6% | 1.8% |
| CatalogBenchmeasured by us | Reliably publish-ready% of products | 0.0% | 0.0% |
| CatalogBenchmeasured by us | Unsupported claimsclaims per product | 7.21 | 7.18 |
| CatalogBenchmeasured by us | Channel compliance% of products | 97.0% | 99.4% |
| CatalogBenchmeasured by us | Conflicts caught% of conflicts | 100.0% | 97.0% |
| CatalogBenchmeasured by us | Content quality% of checks | 96.5% | 96.0% |
| CatalogBenchmeasured by us | Cost per productUS dollars | $0.0355 | $0.0469 |
| CatalogBenchmeasured by us | Decision accuracy% of decisions | 94.8% | 94.8% |
| CatalogBenchmeasured by us | Failed outputs% of products | 0.0% | 0.0% |
| CatalogBenchmeasured by us | Field accuracy% of missing fields | 95.4% | 94.4% |
| CatalogBenchmeasured by us | Invented values% of filled values | 1.5% | 5.7% |
| CatalogBenchmeasured by us | Publish-ready listings% of products | 42.3% | 47.6% |
| CatalogBenchmeasured by us | Reliably publish-ready% of products | 25.0% | 17.9% |
| CatalogBenchmeasured by us | Unsupported claimsclaims per product | 1.64 | 1.39 |
| ROASBenchmeasured by us | Cost of a runUS dollars | $0.52 | $1.01 |
| ROASBenchmeasured by us | Months over budgetmonths of 12 | 0 | 0 |
| ROASBenchmeasured by us | Overall scorescore out of 100 | 43.5 | 44.1 |
| ROASBenchmeasured by us | Return on ad spendprofit per £1 spent | £0.42 | £0.50 |
| Arena (formerly LMArena)reported | Confirmed task successIPS effect estimate | -0.08 | 0.09 |
| Arena (formerly LMArena)reported | Praise over complaintIPS effect estimate | -0.01 | 0.07 |
| Arena (formerly LMArena)reported | SteerabilityIPS effect estimate | -0.04 | -0.01 |
| Arena (formerly LMArena)reported | Tool groundingIPS effect estimate | -0.00 | 0.00 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,472 | 1,472 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,477 | 1,489 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,480 | 1,458 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,509 | 1,533 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,479 | 1,486 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,487 | 1,488 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,480 | 1,475 |
| tau2-benchreported | Consistency (pass^4)% of tasks | 9.3% | 17.5% |
| tau2-benchreported | Task success (pass^1)% of tasks | 26.0% | 37.1% |
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.