Model comparison
Claude 3.5 Haiku vs Gemma 2 27B
Claude 3.5 Haiku and Gemma 2 27B score almost the same on the Noometry Index (29.2 vs 29.4), so choose on price, context window or the category you care about most.
Last verified . 33 shared benchmarks.
GoogleSummary
- They share 33 benchmarks with published results for both. Claude 3.5 Haiku scores higher in 5 categories and Gemma 2 27B in 3 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in math, where Claude 3.5 Haiku leads 14.7 to 10.7.
- The biggest single-benchmark swing is MATH Level 5: 46.4% for Claude 3.5 Haiku and 27.9% for Gemma 2 27B.
- Gemma 2 27B has downloadable open weights; the other is API-only.
Side by side
| Claude 3.5 Haiku | Gemma 2 27B | |
|---|---|---|
| Provider | Anthropic | |
| Noometry Index | 29.2 | 29.4 |
| Released | 2024-10-22 | 2024-06-24 |
| Weights | Proprietary | Open |
| Context window | — | 8K |
| Max output | — | 2K |
| Input $ / M tokens | — | $0.65 |
| Output $ / M tokens | — | $0.65 |
| Results tracked | 49 | 34 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemma 2 27B leads
Claude 3.5 Haiku: 32.9 (#265), Gemma 2 27B: 34.1 (#246)
| Benchmark | Claude 3.5 Haiku | Gemma 2 27B |
|---|---|---|
| BigCodeBench Instruct | 46.1% | 42.8% |
| LiveBench Coding | 51.4% | 36% |
| LMArena Coding | 1286 | 1211 |
| BigCodeBench Complete | 59% | 52.5% |
| Aider Polyglot | 28% | — |
| SciCode | 27.4% | — |
| WeirdML | 30.7% | — |
| CadEval | 32% | — |
Agentic & Tool Use Not comparable
Claude 3.5 Haiku: 28.0 (#95), Gemma 2 27B: —
| Benchmark | Claude 3.5 Haiku | Gemma 2 27B |
|---|---|---|
| BALROG | 19.3% | — |
Reasoning Claude 3.5 Haiku leads
Claude 3.5 Haiku: 17.7 (#290), Gemma 2 27B: 15.3 (#315)
| Benchmark | Claude 3.5 Haiku | Gemma 2 27B |
|---|---|---|
| LiveBench Reasoning | 28.1% | 28.1% |
| LMArena Hard Prompts | 1251 | 1198 |
| DTBench | 56.7% | 48% |
| LiveBench Data Analysis | 48.5% | 47.9% |
| Epoch Capabilities Index | 127.15 | 122.08 |
| LiveBench | 43.5% | 38.2% |
| CritPt | 0% | — |
| LMCA | — | 7.1% |
Math Claude 3.5 Haiku leads
Claude 3.5 Haiku: 14.7 (#300), Gemma 2 27B: 10.7 (#311)
| Benchmark | Claude 3.5 Haiku | Gemma 2 27B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 4.3% | 1.4% |
| LiveBench Math | 35.5% | 26.5% |
| LMArena Math | 1244 | 1212 |
| MATH Level 5 | 46.4% | 27.9% |
| Omni-MATH | 22.4% | — |
| FrontierMath (Feb 2025 set) | 0.3% | — |
Knowledge Too close to call
Claude 3.5 Haiku: 18.7 (#281), Gemma 2 27B: 19.0 (#280)
| Benchmark | Claude 3.5 Haiku | Gemma 2 27B |
|---|---|---|
| GPQA Diamond | 38.1% | 36.5% |
| Confabulations | 36.7% | 27.1% |
| LMArena Expert | 1208 | 1172 |
| MMLU | 74.3% | 75.7% |
| MMLU-Pro | 60.5% | — |
| GPQA (HELM) | 36.3% | — |
Multimodal Not comparable
Claude 3.5 Haiku: 26.8 (#117), Gemma 2 27B: —
| Benchmark | Claude 3.5 Haiku | Gemma 2 27B |
|---|---|---|
| LMArena Vision | 1092 | — |
| GeoBench | 34% | — |
Multilingual Claude 3.5 Haiku leads
Claude 3.5 Haiku: 40.0 (#218), Gemma 2 27B: 38.6 (#226)
| Benchmark | Claude 3.5 Haiku | Gemma 2 27B |
|---|---|---|
| LMArena Non-English | 1238 | 1217 |
| LMArena Chinese | 1229 | 1221 |
| LMArena French | 1264 | 1247 |
| LMArena German | 1237 | 1209 |
| LMArena Japanese | 1175 | 1175 |
| LMArena Korean | 1173 | 1174 |
| LMArena Russian | 1253 | 1234 |
| LMArena Spanish | 1261 | 1228 |
Instruction Following Claude 3.5 Haiku leads
Claude 3.5 Haiku: 62.9 (#234), Gemma 2 27B: 60.5 (#249)
| Benchmark | Claude 3.5 Haiku | Gemma 2 27B |
|---|---|---|
| LiveBench Instruction Following | 61.9% | 58.1% |
| LMArena Instruction Following | 1241 | 1206 |
| IFEval | 79.2% | — |
Long Context Too close to call
Claude 3.5 Haiku: 38.3 (#200), Gemma 2 27B: 37.3 (#218)
| Benchmark | Claude 3.5 Haiku | Gemma 2 27B |
|---|---|---|
| LMArena Longer Query | 1261 | 1231 |
Writing & Preference Gemma 2 27B leads
Claude 3.5 Haiku: 42.7 (#234), Gemma 2 27B: 44.2 (#225)
| Benchmark | Claude 3.5 Haiku | Gemma 2 27B |
|---|---|---|
| LMArena Text | 1255 | 1231 |
| LMArena Creative Writing | 1233 | 1241 |
| LMArena Multi-Turn | 1265 | 1224 |
| LiveBench Language | 35.4% | 32.6% |
| Short-Story Creative Writing | 73.5% | — |
| EQ-Bench Creative Writing | 1146 | — |
| WildBench | 76% | — |
Frequently asked questions
Is Claude 3.5 Haiku better than Gemma 2 27B?
Claude 3.5 Haiku and Gemma 2 27B score almost the same on the Noometry Index (29.2 vs 29.4), so choose on price, context window or the category you care about most.
Is Claude 3.5 Haiku or Gemma 2 27B better for coding?
Gemma 2 27B scores higher on coding benchmarks: 34.1 versus 32.9 in the Noometry coding category.
How many benchmarks do Claude 3.5 Haiku and Gemma 2 27B share?
33 benchmarks have published results for both models. Claude 3.5 Haiku has 49 scored results on Noometry and Gemma 2 27B has 34.