Compare / Claude Haiku 4.5 vs GPT-6 Astra
Claude Haiku 4.5 vs GPT-6 Astra
60 results from 6 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 | GPT-6 Astra |
|---|---|---|---|
| BulletBenchmeasured by us | Cost per gameUS dollars | $0.0318 | $0.28 |
| BulletBenchmeasured by us | Games lost on time% of games | 0.0% | 37.5% |
| BulletBenchmeasured by us | Invalid moves% of moves | 0.9% | 0.0% |
| BulletBenchmeasured by us | Ladder Eloladder Elo | 491 | 839 |
| BulletBenchmeasured by us | Median move timemilliseconds | 1.3 s | 5.6 s |
| BulletBenchmeasured by us | Cost per gameUS dollars | $0.0163 | $0.0943 |
| BulletBenchmeasured by us | Games lost on time% of games | 25.0% | 75.0% |
| BulletBenchmeasured by us | Invalid moves% of moves | 0.3% | 0.0% |
| BulletBenchmeasured by us | Ladder Eloladder Elo | 445 | 323 |
| BulletBenchmeasured by us | Median move timemilliseconds | 1.2 s | 2.3 s |
| CatalogBenchmeasured by us | Channel rules broken% of products | 22.6% | 0.0% |
| CatalogBenchmeasured by us | Content quality% of checks | 94.2% | 99.0% |
| CatalogBenchmeasured by us | Failed outputs% of products | 0.0% | 0.0% |
| CatalogBenchmeasured by us | Missing UK information% of products | 0.0% | 0.0% |
| CatalogBenchmeasured by us | Not findable% of products | 6.5% | 5.4% |
| CatalogBenchmeasured by us | Publish-ready listings% of products | 1.2% | 72.6% |
| CatalogBenchmeasured by us | Reliably publish-ready% of products | 0.0% | 67.9% |
| CatalogBenchmeasured by us | Unsupported claimsclaims per product | 7.30 | 0.07 |
| CatalogBenchmeasured by us | Unsupported claims% of products | 95.8% | 3.0% |
| CatalogBenchmeasured by us | Wrong attributes% of products | 51.2% | 20.2% |
| CatalogBenchmeasured by us | Wrong category or variant% of products | 6.5% | 1.8% |
| CatalogBenchmeasured by us | Channel compliance% of products | 92.3% | 100.0% |
| CatalogBenchmeasured by us | Channel rules broken% of products | 7.7% | 0.0% |
| CatalogBenchmeasured by us | Conflicts caught% of conflicts | 100.0% | 100.0% |
| CatalogBenchmeasured by us | Content quality% of checks | 94.9% | 98.3% |
| CatalogBenchmeasured by us | Cost per productUS dollars | $0.0061 | $0.0371 |
| CatalogBenchmeasured by us | Decision accuracy% of decisions | 89.3% | 98.4% |
| CatalogBenchmeasured by us | Failed outputs% of products | 0.0% | 0.0% |
| CatalogBenchmeasured by us | Field accuracy% of missing fields | 93.7% | 94.2% |
| CatalogBenchmeasured by us | Invented values% of filled values | 6.2% | 1.9% |
| CatalogBenchmeasured by us | Missing UK information% of products | 0.0% | 0.0% |
| CatalogBenchmeasured by us | Not findable% of products | 8.9% | 6.0% |
| CatalogBenchmeasured by us | Publish-ready listings% of products | 7.7% | 77.4% |
| CatalogBenchmeasured by us | Reliably publish-ready% of products | 1.8% | 75.0% |
| CatalogBenchmeasured by us | Unsupported claims% of products | 80.4% | 1.2% |
| CatalogBenchmeasured by us | Unsupported claimsclaims per product | 3.50 | 0.02 |
| CatalogBenchmeasured by us | Wrong attributes% of products | 48.2% | 17.3% |
| CatalogBenchmeasured by us | Wrong category or variant% of products | 7.7% | 1.8% |
| ROASBenchmeasured by us | Audience scorescore out of 100 | 30.9 | 68.6 |
| ROASBenchmeasured by us | Business scorescore out of 100 | 7.3 | 57.5 |
| ROASBenchmeasured by us | Consistency scorescore out of 100 | 8.7 | 44.7 |
| ROASBenchmeasured by us | Contribution profitsimulated US dollars | −$205,001 | $599,139 |
| ROASBenchmeasured by us | Cost of a runUS dollars | $0.0298 | $1.23 |
| ROASBenchmeasured by us | Months over budgetmonths of 12 | 2 | 0 |
| ROASBenchmeasured by us | Overall scorescore out of 100 | 17.5 | 56.6 |
| ROASBenchmeasured by us | Planning scorescore out of 100 | 54.4 | 54.9 |
| ROASBenchmeasured by us | Return on ad spendprofit per $1 spent | −$0.21 | $0.67 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,419 | 1,473 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,418 | 1,465 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,420 | 1,483 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,387 | 1,453 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,452 | 1,510 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,415 | 1,473 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,414 | 1,478 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,399 | 1,469 |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 6.0% | 0.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.39 | 0.38 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 52.2 | 68.9 |
| Vectara Hallucination Leaderboardreported | Answer rate% of documents | 99.5% | 100.0% |
| Vectara Hallucination Leaderboardreported | Hallucination rate% of summaries | 9.8% | 8.7% |
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.