Model comparison
Mixtral 8x22B vs Mixtral 8x7B
Mixtral 8x22B and Mixtral 8x7B score almost the same on the Noometry Index (27.1 vs 27.1), so choose on price, context window or the category you care about most.
Last verified . 30 shared benchmarks.
Mistral AISummary
- They share 30 benchmarks with published results for both. Mixtral 8x22B scores higher in 7 categories and Mixtral 8x7B in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in coding, where Mixtral 8x7B leads 32.8 to 24.2.
- The biggest single-benchmark swing is IFEval: 72.4% for Mixtral 8x22B and 57.5% for Mixtral 8x7B.
- Mixtral 8x7B is cheaper at $0.70 / $0.70 per million input/output tokens, against $2 / $6 for Mixtral 8x22B.
- Mixtral 8x22B accepts more context: 64K tokens versus 32K.
Side by side
| Mixtral 8x22B | Mixtral 8x7B | |
|---|---|---|
| Provider | Mistral AI | Mistral AI |
| Noometry Index | 27.1 | 27.1 |
| Released | 2024-04-17 | 2023-12-11 |
| Weights | Open | Open |
| Context window | 64K | 32K |
| Max output | 64K | 32K |
| Input $ / M tokens | $2 | $0.70 |
| Output $ / M tokens | $6 | $0.70 |
| Results tracked | 34 | 38 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Mixtral 8x7B leads
Mixtral 8x22B: 24.2 (#329), Mixtral 8x7B: 32.8 (#269)
| Benchmark | Mixtral 8x22B | Mixtral 8x7B |
|---|---|---|
| LMArena Coding | 1166 | 1126 |
| HumanEval+ | 72% | 39.6% |
| MBPP+ | 64.3% | 49.7% |
| WeirdML | 3.2% | — |
| BigCodeBench Instruct | 40.6% | — |
| BigCodeBench Complete | 50.2% | — |
Agentic & Tool Use Not comparable
Mixtral 8x22B: 23.1 (#127), Mixtral 8x7B: —
| Benchmark | Mixtral 8x22B | Mixtral 8x7B |
|---|---|---|
| Cybench | 7.5% | — |
Reasoning Mixtral 8x22B leads
Mixtral 8x22B: 19.9 (#248), Mixtral 8x7B: 18.2 (#285)
| Benchmark | Mixtral 8x22B | Mixtral 8x7B |
|---|---|---|
| LMArena Hard Prompts | 1150 | 1115 |
| DTBench | 55.1% | 49.6% |
| Epoch Capabilities Index | 122.03 | 118.47 |
| ForecastBench | 56.3 | 56.3 |
| Adversarial NLI | — | 55.2% |
| HellaSwag | — | 86.7% |
| PIQA | — | 83.6% |
| WinoGrande | — | 77.2% |
Math Mixtral 8x22B leads
Mixtral 8x22B: 22.9 (#275), Mixtral 8x7B: 18.8 (#289)
| Benchmark | Mixtral 8x22B | Mixtral 8x7B |
|---|---|---|
| Omni-MATH | 16.3% | 10.5% |
| LMArena Math | 1184 | 1147 |
| MATH Level 5 | 24.2% | 10% |
| GSM8K | — | 74.4% |
Knowledge Mixtral 8x22B leads
Mixtral 8x22B: 15.1 (#293), Mixtral 8x7B: 11.0 (#301)
| Benchmark | Mixtral 8x22B | Mixtral 8x7B |
|---|---|---|
| GPQA Diamond | 34.1% | 30.6% |
| MMLU-Pro | 46% | 33.5% |
| GPQA (HELM) | 33.4% | 29.6% |
| LMArena Expert | 1113 | 1088 |
| MMLU | 77.8% | 70.6% |
| ARC (AI2) Challenge | — | 87.3% |
| OpenBookQA | — | 85.8% |
| TriviaQA | — | 82.2% |
Multilingual Mixtral 8x22B leads
Mixtral 8x22B: 32.8 (#255), Mixtral 8x7B: 29.6 (#266)
| Benchmark | Mixtral 8x22B | Mixtral 8x7B |
|---|---|---|
| LMArena Non-English | 1128 | 1077 |
| LMArena Chinese | 1116 | 1055 |
| LMArena French | 1166 | 1166 |
| LMArena German | 1141 | 1114 |
| LMArena Japanese | 1037 | 931 |
| LMArena Korean | 1057 | 968 |
| LMArena Russian | 1158 | 1090 |
| LMArena Spanish | 1151 | 1111 |
Instruction Following Mixtral 8x22B leads
Mixtral 8x22B: 57.7 (#266), Mixtral 8x7B: 51.0 (#297)
| Benchmark | Mixtral 8x22B | Mixtral 8x7B |
|---|---|---|
| IFEval | 72.4% | 57.5% |
| LMArena Instruction Following | 1147 | 1109 |
Long Context Mixtral 8x22B leads
Mixtral 8x22B: 34.7 (#247), Mixtral 8x7B: 33.4 (#260)
| Benchmark | Mixtral 8x22B | Mixtral 8x7B |
|---|---|---|
| LMArena Longer Query | 1144 | 1103 |
Writing & Preference Mixtral 8x22B leads
Mixtral 8x22B: 36.9 (#262), Mixtral 8x7B: 34.2 (#270)
| Benchmark | Mixtral 8x22B | Mixtral 8x7B |
|---|---|---|
| LMArena Text | 1162 | 1132 |
| LMArena Creative Writing | 1141 | 1109 |
| WildBench | 71.1% | 67.3% |
| LMArena Multi-Turn | 1130 | 1115 |
Frequently asked questions
Is Mixtral 8x22B better than Mixtral 8x7B?
Mixtral 8x22B and Mixtral 8x7B score almost the same on the Noometry Index (27.1 vs 27.1), so choose on price, context window or the category you care about most.
Which is cheaper, Mixtral 8x22B or Mixtral 8x7B?
Mixtral 8x7B is cheaper. It lists at $0.70 per million input tokens and $0.70 per million output tokens; Mixtral 8x22B lists at $2 and $6.
Is Mixtral 8x22B or Mixtral 8x7B better for coding?
Mixtral 8x7B scores higher on coding benchmarks: 32.8 versus 24.2 in the Noometry coding category.
Which has the bigger context window?
Mixtral 8x22B does, with 64K tokens against 32K.
How many benchmarks do Mixtral 8x22B and Mixtral 8x7B share?
30 benchmarks have published results for both models. Mixtral 8x22B has 34 scored results on Noometry and Mixtral 8x7B has 38.