Model comparison
Codellama 34b Instruct vs Llama 2-13B
Codellama 34b Instruct is the stronger model overall, scoring 30.8 to 29.6 on the Noometry Index.
Last verified . 10 shared benchmarks.
MetaSummary
- They share 10 benchmarks with published results for both. Codellama 34b Instruct scores higher in 1 category and Llama 2-13B in 6 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Codellama 34b Instruct leads 19.6 to 12.8.
Side by side
| Codellama 34b Instruct | Llama 2-13B | |
|---|---|---|
| Provider | Meta | Meta |
| Noometry Index | 30.8 | 29.6 |
| Released | — | 2023-07-18 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 14 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Llama 2-13B leads
Codellama 34b Instruct: 28.5 (#314), Llama 2-13B: 30.9 (#291)
| Benchmark | Codellama 34b Instruct | Llama 2-13B |
|---|---|---|
| LMArena Coding | 1046 | 1062 |
| BigCodeBench Instruct | 29% | — |
| BigCodeBench Complete | 37.1% | — |
| HumanEval+ | 43.9% | — |
| MBPP+ | 56.3% | — |
Reasoning Codellama 34b Instruct leads
Codellama 34b Instruct: 19.6 (#255), Llama 2-13B: 12.8 (#337)
| Benchmark | Codellama 34b Instruct | Llama 2-13B |
|---|---|---|
| LMArena Hard Prompts | 1032 | 1051 |
| Chess Puzzles | — | 0% |
| DTBench | — | 42.2% |
| BIG-Bench Hard | — | 58.2% |
| Epoch Capabilities Index | — | 106.17 |
| HellaSwag | — | 80.7% |
| LAMBADA | — | 76.5% |
| PIQA | — | 80.8% |
| WinoGrande | — | 72.8% |
Math Too close to call
Codellama 34b Instruct: 31.0 (#230), Llama 2-13B: 31.1 (#229)
| Benchmark | Codellama 34b Instruct | Llama 2-13B |
|---|---|---|
| LMArena Math | 1056 | 1065 |
| GSM8K | — | 36.9% |
Knowledge Not comparable
Codellama 34b Instruct: —, Llama 2-13B: 28.1 (#249)
| Benchmark | Codellama 34b Instruct | Llama 2-13B |
|---|---|---|
| LMArena Expert | — | 1030 |
| ARC (AI2) Challenge | — | 60.3% |
| BoolQ | — | 82.4% |
| MMLU | — | 55.6% |
| OpenBookQA | — | 57% |
| TriviaQA | — | 79.6% |
Multimodal Not comparable
Codellama 34b Instruct: —, Llama 2-13B: —
| Benchmark | Codellama 34b Instruct | Llama 2-13B |
|---|---|---|
| ScienceQA | — | 55.8% |
Multilingual Too close to call
Codellama 34b Instruct: 25.8 (#284), Llama 2-13B: 26.5 (#279)
| Benchmark | Codellama 34b Instruct | Llama 2-13B |
|---|---|---|
| LMArena Non-English | 1011 | 1024 |
| LMArena Chinese | 976 | 1001 |
| LMArena French | — | 1044 |
| LMArena German | — | 1009 |
| LMArena Japanese | — | 894 |
| LMArena Korean | — | 953 |
| LMArena Russian | — | 1055 |
| LMArena Spanish | — | 1087 |
Instruction Following Llama 2-13B leads
Codellama 34b Instruct: 52.2 (#291), Llama 2-13B: 53.3 (#287)
| Benchmark | Codellama 34b Instruct | Llama 2-13B |
|---|---|---|
| LMArena Instruction Following | 1028 | 1045 |
Long Context Llama 2-13B leads
Codellama 34b Instruct: 30.9 (#284), Llama 2-13B: 32.3 (#269)
| Benchmark | Codellama 34b Instruct | Llama 2-13B |
|---|---|---|
| LMArena Longer Query | 1013 | 1064 |
Writing & Preference Llama 2-13B leads
Codellama 34b Instruct: 28.2 (#297), Llama 2-13B: 29.8 (#289)
| Benchmark | Codellama 34b Instruct | Llama 2-13B |
|---|---|---|
| LMArena Text | 1066 | 1084 |
| LMArena Creative Writing | 1032 | 1047 |
| LMArena Multi-Turn | 1015 | 1050 |
Frequently asked questions
Is Codellama 34b Instruct better than Llama 2-13B?
Codellama 34b Instruct is the stronger model overall, scoring 30.8 to 29.6 on the Noometry Index.
Is Codellama 34b Instruct or Llama 2-13B better for coding?
Llama 2-13B scores higher on coding benchmarks: 30.9 versus 28.5 in the Noometry coding category.
How many benchmarks do Codellama 34b Instruct and Llama 2-13B share?
10 benchmarks have published results for both models. Codellama 34b Instruct has 14 scored results on Noometry and Llama 2-13B has 32.