Model comparison
Claude 3 Haiku vs Codellama 70b Instruct
Codellama 70b Instruct is the stronger model overall, scoring 33.7 to 25.9 on the Noometry Index.
Last verified . 7 shared benchmarks.
MetaSummary
- They share 7 benchmarks with published results for both. Claude 3 Haiku scores higher in 2 categories and Codellama 70b Instruct in 3 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in multilingual, where Claude 3 Haiku leads 36.0 to 24.8.
- Codellama 70b Instruct has downloadable open weights; the other is API-only.
Side by side
| Claude 3 Haiku | Codellama 70b Instruct | |
|---|---|---|
| Provider | Anthropic | Meta |
| Noometry Index | 25.9 | 33.7 |
| Released | 2024-03-07 | — |
| Weights | Proprietary | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 37 | 7 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Codellama 70b Instruct leads
Claude 3 Haiku: 26.4 (#325), Codellama 70b Instruct: 37.6 (#193)
| Benchmark | Claude 3 Haiku | Codellama 70b Instruct |
|---|---|---|
| BigCodeBench Instruct | 39.4% | 40.7% |
| BigCodeBench Complete | 50.1% | 49.6% |
| HumanEval+ | 68.9% | 65.9% |
| WeirdML | 9.8% | — |
| LMArena Coding | 1199 | — |
| CadEval | 12% | — |
| MBPP+ | 68.8% | — |
Reasoning Codellama 70b Instruct leads
Claude 3 Haiku: 16.3 (#307), Codellama 70b Instruct: 20.1 (#242)
| Benchmark | Claude 3 Haiku | Codellama 70b Instruct |
|---|---|---|
| LMArena Hard Prompts | 1174 | 1052 |
| Kagi LLM Benchmark | 34.2% | — |
| DTBench | 50.1% | — |
| LMCA | 8.8% | — |
| Epoch Capabilities Index | 118.35 | — |
| ForecastBench | 53.2 | — |
| WinoGrande | 74.2% | — |
Math Not comparable
Claude 3 Haiku: 9.8 (#319), Codellama 70b Instruct: —
| Benchmark | Claude 3 Haiku | Codellama 70b Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.8% | — |
| LMArena Math | 1188 | — |
| MATH Level 5 | 14.9% | — |
Knowledge Not comparable
Claude 3 Haiku: 17.3 (#285), Codellama 70b Instruct: —
| Benchmark | Claude 3 Haiku | Codellama 70b Instruct |
|---|---|---|
| GPQA Diamond | 36.3% | — |
| Confabulations | 34.2% | — |
| LMArena Expert | 1148 | — |
| MMLU | 73.8% | — |
Multimodal Not comparable
Claude 3 Haiku: 23.6 (#128), Codellama 70b Instruct: —
| Benchmark | Claude 3 Haiku | Codellama 70b Instruct |
|---|---|---|
| LMArena Vision | 950 | — |
| ScienceQA | 72% | — |
Multilingual Claude 3 Haiku leads
Claude 3 Haiku: 36.0 (#243), Codellama 70b Instruct: 24.8 (#288)
| Benchmark | Claude 3 Haiku | Codellama 70b Instruct |
|---|---|---|
| LMArena Non-English | 1178 | 992 |
| LMArena Chinese | 1155 | — |
| LMArena French | 1195 | — |
| LMArena German | 1174 | — |
| LMArena Japanese | 1102 | — |
| LMArena Korean | 1109 | — |
| LMArena Russian | 1204 | — |
| LMArena Spanish | 1166 | — |
Instruction Following Claude 3 Haiku leads
Claude 3 Haiku: 61.3 (#247), Codellama 70b Instruct: 51.9 (#293)
| Benchmark | Claude 3 Haiku | Codellama 70b Instruct |
|---|---|---|
| LMArena Instruction Following | 1173 | 1024 |
Long Context Not comparable
Claude 3 Haiku: 36.1 (#237), Codellama 70b Instruct: —
| Benchmark | Claude 3 Haiku | Codellama 70b Instruct |
|---|---|---|
| LMArena Longer Query | 1190 | — |
Writing & Preference Codellama 70b Instruct leads
Claude 3 Haiku: 29.7 (#291), Codellama 70b Instruct: 33.4 (#277)
| Benchmark | Claude 3 Haiku | Codellama 70b Instruct |
|---|---|---|
| LMArena Text | 1195 | 1057 |
| LMArena Creative Writing | 1157 | — |
| EQ-Bench Creative Writing | 717 | — |
| LMArena Multi-Turn | 1190 | — |
Frequently asked questions
Is Claude 3 Haiku better than Codellama 70b Instruct?
Codellama 70b Instruct is the stronger model overall, scoring 33.7 to 25.9 on the Noometry Index.
Is Claude 3 Haiku or Codellama 70b Instruct better for coding?
Codellama 70b Instruct scores higher on coding benchmarks: 37.6 versus 26.4 in the Noometry coding category.
How many benchmarks do Claude 3 Haiku and Codellama 70b Instruct share?
7 benchmarks have published results for both models. Claude 3 Haiku has 37 scored results on Noometry and Codellama 70b Instruct has 7.