Model comparison
Claude 3.5 Haiku vs Deepseek Coder v2
Deepseek Coder v2 is the stronger model overall, scoring 35.9 to 29.2 on the Noometry Index.
Last verified . 19 shared benchmarks.
DeepSeekSummary
- They share 19 benchmarks with published results for both. Claude 3.5 Haiku scores higher in 4 categories and Deepseek Coder v2 in 4 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Deepseek Coder v2 leads 34.9 to 14.7.
- Deepseek Coder v2 has downloadable open weights; the other is API-only.
Side by side
| Claude 3.5 Haiku | Deepseek Coder v2 | |
|---|---|---|
| Provider | Anthropic | DeepSeek |
| Noometry Index | 29.2 | 35.9 |
| Released | 2024-10-22 | 2024-06-17 |
| Weights | Proprietary | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 49 | 24 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Deepseek Coder v2 leads
Claude 3.5 Haiku: 32.9 (#265), Deepseek Coder v2: 38.1 (#183)
| Benchmark | Claude 3.5 Haiku | Deepseek Coder v2 |
|---|---|---|
| BigCodeBench Instruct | 46.1% | 48.2% |
| LMArena Coding | 1286 | 1251 |
| BigCodeBench Complete | 59% | 59.7% |
| Aider Polyglot | 28% | — |
| SciCode | 27.4% | — |
| WeirdML | 30.7% | — |
| LiveBench Coding | 51.4% | — |
| CadEval | 32% | — |
| HumanEval+ | — | 82.3% |
| MBPP+ | — | 75.1% |
Agentic & Tool Use Not comparable
Claude 3.5 Haiku: 28.0 (#95), Deepseek Coder v2: —
| Benchmark | Claude 3.5 Haiku | Deepseek Coder v2 |
|---|---|---|
| BALROG | 19.3% | — |
Reasoning Deepseek Coder v2 leads
Claude 3.5 Haiku: 17.7 (#290), Deepseek Coder v2: 23.6 (#176)
| Benchmark | Claude 3.5 Haiku | Deepseek Coder v2 |
|---|---|---|
| LMArena Hard Prompts | 1251 | 1207 |
| CritPt | 0% | — |
| LiveBench Reasoning | 28.1% | — |
| DTBench | 56.7% | — |
| LiveBench Data Analysis | 48.5% | — |
| Epoch Capabilities Index | 127.15 | — |
| LiveBench | 43.5% | — |
| WinoGrande | — | 83.7% |
Math Deepseek Coder v2 leads
Claude 3.5 Haiku: 14.7 (#300), Deepseek Coder v2: 34.9 (#190)
| Benchmark | Claude 3.5 Haiku | Deepseek Coder v2 |
|---|---|---|
| LMArena Math | 1244 | 1241 |
| OTIS Mock AIME 2024-2025 | 4.3% | — |
| Omni-MATH | 22.4% | — |
| LiveBench Math | 35.5% | — |
| MATH Level 5 | 46.4% | — |
| FrontierMath (Feb 2025 set) | 0.3% | — |
| GSM8K | — | 94.5% |
Knowledge Deepseek Coder v2 leads
Claude 3.5 Haiku: 18.7 (#281), Deepseek Coder v2: 32.3 (#212)
| Benchmark | Claude 3.5 Haiku | Deepseek Coder v2 |
|---|---|---|
| LMArena Expert | 1208 | 1181 |
| GPQA Diamond | 38.1% | — |
| MMLU-Pro | 60.5% | — |
| Confabulations | 36.7% | — |
| GPQA (HELM) | 36.3% | — |
| ARC (AI2) Challenge | — | 64.3% |
| MMLU | 74.3% | — |
Multimodal Not comparable
Claude 3.5 Haiku: 26.8 (#117), Deepseek Coder v2: —
| Benchmark | Claude 3.5 Haiku | Deepseek Coder v2 |
|---|---|---|
| LMArena Vision | 1092 | — |
| GeoBench | 34% | — |
Multilingual Claude 3.5 Haiku leads
Claude 3.5 Haiku: 40.0 (#218), Deepseek Coder v2: 36.3 (#240)
| Benchmark | Claude 3.5 Haiku | Deepseek Coder v2 |
|---|---|---|
| LMArena Non-English | 1238 | 1182 |
| LMArena Chinese | 1229 | 1201 |
| LMArena French | 1264 | 1185 |
| LMArena German | 1237 | 1164 |
| LMArena Japanese | 1175 | 1126 |
| LMArena Korean | 1173 | 1104 |
| LMArena Russian | 1253 | 1188 |
| LMArena Spanish | 1261 | 1153 |
Instruction Following Claude 3.5 Haiku leads
Claude 3.5 Haiku: 62.9 (#234), Deepseek Coder v2: 61.7 (#242)
| Benchmark | Claude 3.5 Haiku | Deepseek Coder v2 |
|---|---|---|
| LMArena Instruction Following | 1241 | 1180 |
| LiveBench Instruction Following | 61.9% | — |
| IFEval | 79.2% | — |
Long Context Claude 3.5 Haiku leads
Claude 3.5 Haiku: 38.3 (#200), Deepseek Coder v2: 37.0 (#224)
| Benchmark | Claude 3.5 Haiku | Deepseek Coder v2 |
|---|---|---|
| LMArena Longer Query | 1261 | 1219 |
Writing & Preference Claude 3.5 Haiku leads
Claude 3.5 Haiku: 42.7 (#234), Deepseek Coder v2: 38.2 (#253)
| Benchmark | Claude 3.5 Haiku | Deepseek Coder v2 |
|---|---|---|
| LMArena Text | 1255 | 1191 |
| LMArena Creative Writing | 1233 | 1120 |
| LMArena Multi-Turn | 1265 | 1177 |
| Short-Story Creative Writing | 73.5% | — |
| EQ-Bench Creative Writing | 1146 | — |
| WildBench | 76% | — |
| LiveBench Language | 35.4% | — |
Frequently asked questions
Is Claude 3.5 Haiku better than Deepseek Coder v2?
Deepseek Coder v2 is the stronger model overall, scoring 35.9 to 29.2 on the Noometry Index.
Is Claude 3.5 Haiku or Deepseek Coder v2 better for coding?
Deepseek Coder v2 scores higher on coding benchmarks: 38.1 versus 32.9 in the Noometry coding category.
How many benchmarks do Claude 3.5 Haiku and Deepseek Coder v2 share?
19 benchmarks have published results for both models. Claude 3.5 Haiku has 49 scored results on Noometry and Deepseek Coder v2 has 24.