Anthropic
Claude 3 Haiku benchmark results
9 results on 9 benchmarks from 5 sources.
Closed Released: Mar 7, 2024 Updated Oct 22, 2026
Also listed as: claude-3-haiku-20240307 · Claude 3 Haiku (20240307) · Claude 3.5 Haiku · Tools + Claude 3.5 Haiku · Claude-3-Haiku · Claude-3-Haiku-20240307
All benchmark results
| Benchmarks | Score | Rank | Setting | Source type | Date |
|---|---|---|---|---|---|
| LMArena Text | 1262 | 315 | Overall, style control on; 117701 votes; 95% CI 1258–1265“claude-3-haiku-20240307” | Official leaderboard | Oct 8, 2026 |
| LMArena Vision | 1000 | 149 | Overall, style control on; 13380 votes; 95% CI 988–1012“claude-3-haiku-20240307” | Official leaderboard | Oct 2, 2026 |
| LiveCodeBench | 23.9% | 21 | Code generation, 182 problems dated 2025-01-01…2025-04-07“Claude-3-Haiku” | Official leaderboard | Apr 7, 2025 |
| GPQA Diamond | 36.3% | 246 | Epoch AI run, accuracy | Independent run | Jan 27, 2025 |
| MMLU | 73.8% | 45 | Exact match; 5-shot | Official leaderboard via Epoch AI | Mar 7, 2024 |
| MMLU-Pro | 42.3% | 185 | Overall accuracy; data source: TIGER-Lab“Claude-3-Haiku-20240307” | Official leaderboard | Mar 11, 2026 |
| SWE-bench Verified | 40.6% | 82 | agent: Tools“claude-3-haiku-20240307” | Official leaderboard | Oct 22, 2024 |
| AIME (OTIS Mock AIME 2024–2025) | 1.8% | 252 | Epoch AI run, OTIS Mock AIME 2024–2025 | Independent run | Feb 25, 2025 |
| MATH Level 5 | 14.9% | 92 | Epoch AI run, accuracy | Independent run | Jan 27, 2025 |
FAQ
Where do these Claude 3 Haiku scores come from?
From Epoch AI Benchmarking Hub, LiveCodeBench leaderboard, LMArena leaderboard dataset, MMLU-Pro leaderboard (TIGER-Lab), and SWE-bench leaderboards (swebench.com). Each row links to the original leaderboard, dataset or report, and vendor-reported numbers are labelled.