Model comparison
DeepSeek LLM 67B vs Mixtral 8x7B
Mixtral 8x7B is the stronger model overall, scoring 27.1 to 24.9 on the Noometry Index.
Last verified . 13 shared benchmarks.
DeepSeek
Mistral AISummary
- They share 13 benchmarks with published results for both. DeepSeek LLM 67B scores higher in 1 category and Mixtral 8x7B in 7 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in math, where Mixtral 8x7B leads 18.8 to 8.7.
- The biggest single-benchmark swing is GPQA Diamond: 24.6% for DeepSeek LLM 67B and 30.6% for Mixtral 8x7B.
Side by side
| DeepSeek LLM 67B | Mixtral 8x7B | |
|---|---|---|
| Provider | DeepSeek | Mistral AI |
| Noometry Index | 24.9 | 27.1 |
| Released | 2023-11-29 | 2023-12-11 |
| Weights | Open | Open |
| Context window | — | 32K |
| Max output | — | 32K |
| Input $ / M tokens | — | $0.70 |
| Output $ / M tokens | — | $0.70 |
| Results tracked | 15 | 38 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
DeepSeek LLM 67B: 31.9 (#278), Mixtral 8x7B: 32.8 (#269)
| Benchmark | DeepSeek LLM 67B | Mixtral 8x7B |
|---|---|---|
| LMArena Coding | 1096 | 1126 |
| HumanEval+ | — | 39.6% |
| MBPP+ | — | 49.7% |
Reasoning Mixtral 8x7B leads
DeepSeek LLM 67B: 16.5 (#304), Mixtral 8x7B: 18.2 (#285)
| Benchmark | DeepSeek LLM 67B | Mixtral 8x7B |
|---|---|---|
| LMArena Hard Prompts | 1070 | 1115 |
| Epoch Capabilities Index | 110.5 | 118.47 |
| Chess Puzzles | 0% | — |
| DTBench | — | 49.6% |
| Adversarial NLI | — | 55.2% |
| ForecastBench | — | 56.3 |
| HellaSwag | — | 86.7% |
| PIQA | — | 83.6% |
| WinoGrande | — | 77.2% |
Math Mixtral 8x7B leads
DeepSeek LLM 67B: 8.7 (#324), Mixtral 8x7B: 18.8 (#289)
| Benchmark | DeepSeek LLM 67B | Mixtral 8x7B |
|---|---|---|
| LMArena Math | 1108 | 1147 |
| MATH Level 5 | 6.4% | 10% |
| OTIS Mock AIME 2024-2025 | 0.8% | — |
| Omni-MATH | — | 10.5% |
| GSM8K | — | 74.4% |
Knowledge Mixtral 8x7B leads
DeepSeek LLM 67B: 7.0 (#313), Mixtral 8x7B: 11.0 (#301)
| Benchmark | DeepSeek LLM 67B | Mixtral 8x7B |
|---|---|---|
| GPQA Diamond | 24.6% | 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 Too close to call
DeepSeek LLM 67B: 29.4 (#267), Mixtral 8x7B: 29.6 (#266)
| Benchmark | DeepSeek LLM 67B | Mixtral 8x7B |
|---|---|---|
| LMArena Non-English | 1073 | 1077 |
| LMArena Chinese | 1132 | 1055 |
| LMArena French | — | 1166 |
| LMArena German | — | 1114 |
| LMArena Japanese | — | 931 |
| LMArena Korean | — | 968 |
| LMArena Russian | — | 1090 |
| LMArena Spanish | — | 1111 |
Instruction Following DeepSeek LLM 67B leads
DeepSeek LLM 67B: 55.4 (#277), Mixtral 8x7B: 51.0 (#297)
| Benchmark | DeepSeek LLM 67B | Mixtral 8x7B |
|---|---|---|
| LMArena Instruction Following | 1079 | 1109 |
| IFEval | — | 57.5% |
Long Context Too close to call
DeepSeek LLM 67B: 33.1 (#265), Mixtral 8x7B: 33.4 (#260)
| Benchmark | DeepSeek LLM 67B | Mixtral 8x7B |
|---|---|---|
| LMArena Longer Query | 1092 | 1103 |
Writing & Preference Mixtral 8x7B leads
DeepSeek LLM 67B: 31.6 (#282), Mixtral 8x7B: 34.2 (#270)
| Benchmark | DeepSeek LLM 67B | Mixtral 8x7B |
|---|---|---|
| LMArena Text | 1105 | 1132 |
| LMArena Creative Writing | 1067 | 1109 |
| LMArena Multi-Turn | 1082 | 1115 |
| WildBench | — | 67.3% |
Frequently asked questions
Is DeepSeek LLM 67B better than Mixtral 8x7B?
Mixtral 8x7B is the stronger model overall, scoring 27.1 to 24.9 on the Noometry Index.
Is DeepSeek LLM 67B or Mixtral 8x7B better for coding?
They score almost the same on coding (31.9 vs 32.8); test both on your own repository before choosing.
How many benchmarks do DeepSeek LLM 67B and Mixtral 8x7B share?
13 benchmarks have published results for both models. DeepSeek LLM 67B has 15 scored results on Noometry and Mixtral 8x7B has 38.