Model comparison
Claude 2 vs Claude 3.5 Haiku
Claude 3.5 Haiku is the stronger model overall, scoring 29.2 to 25.0 on the Noometry Index.
Last verified . 6 shared benchmarks.
Summary
- They share 6 benchmarks with published results for both. Claude 2 scores higher in 1 category and Claude 3.5 Haiku in 2 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in math, where Claude 3.5 Haiku leads 14.7 to 9.3.
- The biggest single-benchmark swing is MATH Level 5: 11.7% for Claude 2 and 46.4% for Claude 3.5 Haiku.
Side by side
| Claude 2 | Claude 3.5 Haiku | |
|---|---|---|
| Provider | Anthropic | Anthropic |
| Noometry Index | 25.0 | 29.2 |
| Released | 2023-07-11 | 2024-10-22 |
| Weights | Proprietary | Proprietary |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 8 | 49 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
Claude 2: —, Claude 3.5 Haiku: 32.9 (#265)
| Benchmark | Claude 2 | Claude 3.5 Haiku |
|---|---|---|
| Aider Polyglot | — | 28% |
| SciCode | — | 27.4% |
| WeirdML | — | 30.7% |
| BigCodeBench Instruct | — | 46.1% |
| LiveBench Coding | — | 51.4% |
| LMArena Coding | — | 1286 |
| BigCodeBench Complete | — | 59% |
| CadEval | — | 32% |
| HumanEval+ | 61.6% | — |
Agentic & Tool Use Not comparable
Claude 2: —, Claude 3.5 Haiku: 28.0 (#95)
| Benchmark | Claude 2 | Claude 3.5 Haiku |
|---|---|---|
| BALROG | — | 19.3% |
Reasoning Claude 2 leads
Claude 2: 21.7 (#216), Claude 3.5 Haiku: 17.7 (#290)
| Benchmark | Claude 2 | Claude 3.5 Haiku |
|---|---|---|
| DTBench | 51.9% | 56.7% |
| Epoch Capabilities Index | 120.13 | 127.15 |
| CritPt | — | 0% |
| LiveBench Reasoning | — | 28.1% |
| LMArena Hard Prompts | — | 1251 |
| LiveBench Data Analysis | — | 48.5% |
| LiveBench | — | 43.5% |
Math Claude 3.5 Haiku leads
Claude 2: 9.3 (#320), Claude 3.5 Haiku: 14.7 (#300)
| Benchmark | Claude 2 | Claude 3.5 Haiku |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 2.5% | 4.3% |
| MATH Level 5 | 11.7% | 46.4% |
| Omni-MATH | — | 22.4% |
| LiveBench Math | — | 35.5% |
| LMArena Math | — | 1244 |
| FrontierMath (Feb 2025 set) | — | 0.3% |
Knowledge Claude 3.5 Haiku leads
Claude 2: 16.9 (#287), Claude 3.5 Haiku: 18.7 (#281)
| Benchmark | Claude 2 | Claude 3.5 Haiku |
|---|---|---|
| GPQA Diamond | 34.7% | 38.1% |
| MMLU | 78.5% | 74.3% |
| MMLU-Pro | — | 60.5% |
| Confabulations | — | 36.7% |
| GPQA (HELM) | — | 36.3% |
| LMArena Expert | — | 1208 |
| TriviaQA | 87.5% | — |
Multimodal Not comparable
Claude 2: —, Claude 3.5 Haiku: 26.8 (#117)
| Benchmark | Claude 2 | Claude 3.5 Haiku |
|---|---|---|
| LMArena Vision | — | 1092 |
| GeoBench | — | 34% |
Multilingual Not comparable
Claude 2: —, Claude 3.5 Haiku: 40.0 (#218)
| Benchmark | Claude 2 | Claude 3.5 Haiku |
|---|---|---|
| LMArena Non-English | — | 1238 |
| LMArena Chinese | — | 1229 |
| LMArena French | — | 1264 |
| LMArena German | — | 1237 |
| LMArena Japanese | — | 1175 |
| LMArena Korean | — | 1173 |
| LMArena Russian | — | 1253 |
| LMArena Spanish | — | 1261 |
Instruction Following Not comparable
Claude 2: —, Claude 3.5 Haiku: 62.9 (#234)
| Benchmark | Claude 2 | Claude 3.5 Haiku |
|---|---|---|
| LiveBench Instruction Following | — | 61.9% |
| IFEval | — | 79.2% |
| LMArena Instruction Following | — | 1241 |
Long Context Not comparable
Claude 2: —, Claude 3.5 Haiku: 38.3 (#200)
| Benchmark | Claude 2 | Claude 3.5 Haiku |
|---|---|---|
| LMArena Longer Query | — | 1261 |
Writing & Preference Not comparable
Claude 2: —, Claude 3.5 Haiku: 42.7 (#234)
| Benchmark | Claude 2 | Claude 3.5 Haiku |
|---|---|---|
| LMArena Text | — | 1255 |
| LMArena Creative Writing | — | 1233 |
| Short-Story Creative Writing | — | 73.5% |
| EQ-Bench Creative Writing | — | 1146 |
| WildBench | — | 76% |
| LMArena Multi-Turn | — | 1265 |
| LiveBench Language | — | 35.4% |
Frequently asked questions
Is Claude 2 better than Claude 3.5 Haiku?
Claude 3.5 Haiku is the stronger model overall, scoring 29.2 to 25.0 on the Noometry Index.
How many benchmarks do Claude 2 and Claude 3.5 Haiku share?
6 benchmarks have published results for both models. Claude 2 has 8 scored results on Noometry and Claude 3.5 Haiku has 49.