Compare / Claude Haiku 4.5 vs Kimi K3
Claude Haiku 4.5 vs Kimi K3
49 results from 4 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | Claude Haiku 4.5 | Kimi K3 |
|---|---|---|---|
| BulletBenchmeasured by us | Cost per gameUS dollars | $0.0163 | $0.0540 |
| BulletBenchmeasured by us | Games lost on time% of games | 25.0% | 91.7% |
| BulletBenchmeasured by us | Invalid moves% of moves | 0.3% | 0.0% |
| BulletBenchmeasured by us | Ladder Eloladder Elo | 445 | 33 |
| BulletBenchmeasured by us | Median move timemilliseconds | 1.2 s | 2.6 s |
| CatalogBenchmeasured by us | Channel rules broken% of products | 22.6% | 0.6% |
| CatalogBenchmeasured by us | Content quality% of checks | 94.2% | 96.5% |
| CatalogBenchmeasured by us | Failed outputs% of products | 0.0% | 0.6% |
| CatalogBenchmeasured by us | Missing UK information% of products | 0.0% | 0.0% |
| CatalogBenchmeasured by us | Not findable% of products | 6.5% | 3.0% |
| CatalogBenchmeasured by us | Publish-ready listings% of products | 1.2% | 16.1% |
| CatalogBenchmeasured by us | Reliably publish-ready% of products | 0.0% | 1.8% |
| CatalogBenchmeasured by us | Unsupported claimsclaims per product | 7.30 | 3.45 |
| CatalogBenchmeasured by us | Unsupported claims% of products | 95.8% | 81.0% |
| CatalogBenchmeasured by us | Wrong attributes% of products | 51.2% | 20.8% |
| CatalogBenchmeasured by us | Wrong category or variant% of products | 6.5% | 4.2% |
| CatalogBenchmeasured by us | Channel compliance% of products | 92.3% | 99.4% |
| CatalogBenchmeasured by us | Channel rules broken% of products | 7.7% | 0.6% |
| CatalogBenchmeasured by us | Conflicts caught% of conflicts | 100.0% | 97.0% |
| CatalogBenchmeasured by us | Content quality% of checks | 94.9% | 96.9% |
| CatalogBenchmeasured by us | Cost per productUS dollars | $0.0061 | $0.0469 |
| CatalogBenchmeasured by us | Decision accuracy% of decisions | 89.3% | 96.5% |
| CatalogBenchmeasured by us | Failed outputs% of products | 0.0% | 0.0% |
| CatalogBenchmeasured by us | Field accuracy% of missing fields | 93.7% | 94.4% |
| CatalogBenchmeasured by us | Invented values% of filled values | 6.2% | 5.7% |
| CatalogBenchmeasured by us | Missing UK information% of products | 0.0% | 0.0% |
| CatalogBenchmeasured by us | Not findable% of products | 8.9% | 4.2% |
| CatalogBenchmeasured by us | Publish-ready listings% of products | 7.7% | 61.9% |
| CatalogBenchmeasured by us | Reliably publish-ready% of products | 1.8% | 41.1% |
| CatalogBenchmeasured by us | Unsupported claims% of products | 80.4% | 15.5% |
| CatalogBenchmeasured by us | Unsupported claimsclaims per product | 3.50 | 0.82 |
| CatalogBenchmeasured by us | Wrong attributes% of products | 48.2% | 24.4% |
| CatalogBenchmeasured by us | Wrong category or variant% of products | 7.7% | 3.0% |
| ROASBenchmeasured by us | Audience scorescore out of 100 | 30.9 | 65.1 |
| ROASBenchmeasured by us | Business scorescore out of 100 | 7.3 | 40.2 |
| ROASBenchmeasured by us | Consistency scorescore out of 100 | 8.7 | 28.3 |
| ROASBenchmeasured by us | Contribution profitsimulated US dollars | −$205,001 | $654,270 |
| ROASBenchmeasured by us | Cost of a runUS dollars | $0.0298 | $1.01 |
| ROASBenchmeasured by us | Months over budgetmonths of 12 | 2 | 0 |
| ROASBenchmeasured by us | Overall scorescore out of 100 | 17.5 | 44.1 |
| ROASBenchmeasured by us | Planning scorescore out of 100 | 54.4 | 54.7 |
| ROASBenchmeasured by us | Return on ad spendprofit per $1 spent | −$0.21 | $0.50 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,418 | 1,472 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,420 | 1,489 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,387 | 1,458 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,452 | 1,533 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,415 | 1,486 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,414 | 1,488 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,399 | 1,475 |
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.