Model comparison
Codellama 34b Instruct vs Mixtral 8x7B
Codellama 34b Instruct is the stronger model overall, scoring 30.8 to 27.1 on the Noometry Index.
Last verified . 12 shared benchmarks.
Meta
Mistral AISummary
- They share 12 benchmarks with published results for both. Codellama 34b Instruct scores higher in 3 categories and Mixtral 8x7B in 4 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Codellama 34b Instruct leads 31.0 to 18.8.
Side by side
| Codellama 34b Instruct | Mixtral 8x7B | |
|---|---|---|
| Provider | Meta | Mistral AI |
| Noometry Index | 30.8 | 27.1 |
| Released | — | 2023-12-11 |
| Weights | Open | Open |
| Context window | — | 32K |
| Max output | — | 32K |
| Input $ / M tokens | — | $0.70 |
| Output $ / M tokens | — | $0.70 |
| Results tracked | 14 | 38 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Mixtral 8x7B leads
Codellama 34b Instruct: 28.5 (#314), Mixtral 8x7B: 32.8 (#269)
| Benchmark | Codellama 34b Instruct | Mixtral 8x7B |
|---|---|---|
| LMArena Coding | 1046 | 1126 |
| HumanEval+ | 43.9% | 39.6% |
| MBPP+ | 56.3% | 49.7% |
| BigCodeBench Instruct | 29% | — |
| BigCodeBench Complete | 37.1% | — |
Reasoning Codellama 34b Instruct leads
Codellama 34b Instruct: 19.6 (#255), Mixtral 8x7B: 18.2 (#285)
| Benchmark | Codellama 34b Instruct | Mixtral 8x7B |
|---|---|---|
| LMArena Hard Prompts | 1032 | 1115 |
| DTBench | — | 49.6% |
| Adversarial NLI | — | 55.2% |
| Epoch Capabilities Index | — | 118.47 |
| ForecastBench | — | 56.3 |
| HellaSwag | — | 86.7% |
| PIQA | — | 83.6% |
| WinoGrande | — | 77.2% |
Math Codellama 34b Instruct leads
Codellama 34b Instruct: 31.0 (#230), Mixtral 8x7B: 18.8 (#289)
| Benchmark | Codellama 34b Instruct | Mixtral 8x7B |
|---|---|---|
| LMArena Math | 1056 | 1147 |
| Omni-MATH | — | 10.5% |
| MATH Level 5 | — | 10% |
| GSM8K | — | 74.4% |
Knowledge Not comparable
Codellama 34b Instruct: —, Mixtral 8x7B: 11.0 (#301)
| Benchmark | Codellama 34b Instruct | Mixtral 8x7B |
|---|---|---|
| GPQA Diamond | — | 30.6% |
| MMLU-Pro | — | 33.5% |
| GPQA (HELM) | — | 29.6% |
| LMArena Expert | — | 1088 |
| ARC (AI2) Challenge | — | 87.3% |
| MMLU | — | 70.6% |
| OpenBookQA | — | 85.8% |
| TriviaQA | — | 82.2% |
Multilingual Mixtral 8x7B leads
Codellama 34b Instruct: 25.8 (#284), Mixtral 8x7B: 29.6 (#266)
| Benchmark | Codellama 34b Instruct | Mixtral 8x7B |
|---|---|---|
| LMArena Non-English | 1011 | 1077 |
| LMArena Chinese | 976 | 1055 |
| LMArena French | — | 1166 |
| LMArena German | — | 1114 |
| LMArena Japanese | — | 931 |
| LMArena Korean | — | 968 |
| LMArena Russian | — | 1090 |
| LMArena Spanish | — | 1111 |
Instruction Following Codellama 34b Instruct leads
Codellama 34b Instruct: 52.2 (#291), Mixtral 8x7B: 51.0 (#297)
| Benchmark | Codellama 34b Instruct | Mixtral 8x7B |
|---|---|---|
| LMArena Instruction Following | 1028 | 1109 |
| IFEval | — | 57.5% |
Long Context Mixtral 8x7B leads
Codellama 34b Instruct: 30.9 (#284), Mixtral 8x7B: 33.4 (#260)
| Benchmark | Codellama 34b Instruct | Mixtral 8x7B |
|---|---|---|
| LMArena Longer Query | 1013 | 1103 |
Writing & Preference Mixtral 8x7B leads
Codellama 34b Instruct: 28.2 (#297), Mixtral 8x7B: 34.2 (#270)
| Benchmark | Codellama 34b Instruct | Mixtral 8x7B |
|---|---|---|
| LMArena Text | 1066 | 1132 |
| LMArena Creative Writing | 1032 | 1109 |
| LMArena Multi-Turn | 1015 | 1115 |
| WildBench | — | 67.3% |
Frequently asked questions
Is Codellama 34b Instruct better than Mixtral 8x7B?
Codellama 34b Instruct is the stronger model overall, scoring 30.8 to 27.1 on the Noometry Index.
Is Codellama 34b Instruct or Mixtral 8x7B better for coding?
Mixtral 8x7B scores higher on coding benchmarks: 32.8 versus 28.5 in the Noometry coding category.
How many benchmarks do Codellama 34b Instruct and Mixtral 8x7B share?
12 benchmarks have published results for both models. Codellama 34b Instruct has 14 scored results on Noometry and Mixtral 8x7B has 38.