Claude 3.5 Haiku (20241022): benchmark results
October 22, 2024 Claude 3.5 Haiku snapshot, tracked when sources report the dated API model. Provider: Anthropic. Released 2024-10-22. Access: API.
Unified ELO 1489 ± 1, rank #738 of 1392 rated models, from 228 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Enkrypt AI - Jailbreak Risk | 1.12 | Risk Score | 96.8 |
| Enkrypt AI - Bias Risk | 43.41 | Risk Score | 95.8 |
| Enkrypt AI - Safety Risk | 15.01 | Risk Score | 95.8 |
| EuroEval Lithuanian NLU - ScaLA LT | 50.62 | Linguistic acceptability Score (%) | 95.6 |
| EuroEval Faroese NLU - FoSent | 65.97 | Sentiment classification Score (%) | 92.2 |
| EuroEval Finnish NLU - Scandisent FI | 92.47 | Sentiment classification Score (%) | 92.2 |
| AILuminate Safety | 4 | Safety Grade (1-5) | 91.9 |
| EuroEval Polish NLU - ScaLA PL | 53.53 | Linguistic acceptability Score (%) | 89.3 |
| Enkrypt AI - Risk Score | 21.01 | Risk Score | 86.7 |
| Enkrypt AI - Toxicity Risk | 0.64 | Risk Score | 86.6 |
| UGI - Natural Intelligence | 37.98 | NatInt Score | 85.8 |
| EuroEval Latvian NLU - Latvian Twitter Sentiment | 49.13 | Sentiment classification Score (%) | 84.9 |
Interactive version: theaggregate.ai/model?slug=claude-3-5-haiku-20241022 · How It Works · Data refreshed daily, snapshot 2026-09-05.