Model comparison
DBRX vs Mixtral 8x22B
DBRX is the stronger model overall, scoring 29.4 to 27.1 on the Noometry Index.
Last verified . 21 shared benchmarks.
Databricks
Mistral AISummary
- They share 21 benchmarks with published results for both. DBRX scores higher in 3 categories and Mixtral 8x22B in 5 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in coding, where DBRX leads 32.9 to 24.2.
- The biggest single-benchmark swing is MATH Level 5: 11.7% for DBRX and 24.2% for Mixtral 8x22B.
Side by side
| DBRX | Mixtral 8x22B | |
|---|---|---|
| Provider | Databricks | Mistral AI |
| Noometry Index | 29.4 | 27.1 |
| Released | 2024-03-27 | 2024-04-17 |
| Weights | Open | Open |
| Context window | — | 64K |
| Max output | — | 64K |
| Input $ / M tokens | — | $2 |
| Output $ / M tokens | — | $6 |
| Results tracked | 21 | 34 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DBRX leads
DBRX: 32.9 (#266), Mixtral 8x22B: 24.2 (#329)
| Benchmark | DBRX | Mixtral 8x22B |
|---|---|---|
| LMArena Coding | 1132 | 1166 |
| HumanEval+ | 70.1% | 72% |
| MBPP+ | 55.8% | 64.3% |
| WeirdML | — | 3.2% |
| BigCodeBench Instruct | — | 40.6% |
| BigCodeBench Complete | — | 50.2% |
Agentic & Tool Use Not comparable
DBRX: —, Mixtral 8x22B: 23.1 (#127)
| Benchmark | DBRX | Mixtral 8x22B |
|---|---|---|
| Cybench | — | 7.5% |
Reasoning DBRX leads
DBRX: 21.4 (#222), Mixtral 8x22B: 19.9 (#248)
| Benchmark | DBRX | Mixtral 8x22B |
|---|---|---|
| LMArena Hard Prompts | 1113 | 1150 |
| DTBench | — | 55.1% |
| Epoch Capabilities Index | — | 122.03 |
| ForecastBench | — | 56.3 |
Math DBRX leads
DBRX: 24.3 (#269), Mixtral 8x22B: 22.9 (#275)
| Benchmark | DBRX | Mixtral 8x22B |
|---|---|---|
| LMArena Math | 1145 | 1184 |
| MATH Level 5 | 11.7% | 24.2% |
| Omni-MATH | — | 16.3% |
Knowledge Too close to call
DBRX: 14.9 (#294), Mixtral 8x22B: 15.1 (#293)
| Benchmark | DBRX | Mixtral 8x22B |
|---|---|---|
| GPQA Diamond | 32.9% | 34.1% |
| LMArena Expert | 1076 | 1113 |
| MMLU-Pro | — | 46% |
| GPQA (HELM) | — | 33.4% |
| MMLU | — | 77.8% |
Multilingual Mixtral 8x22B leads
DBRX: 29.3 (#268), Mixtral 8x22B: 32.8 (#255)
| Benchmark | DBRX | Mixtral 8x22B |
|---|---|---|
| LMArena Non-English | 1071 | 1128 |
| LMArena Chinese | 1068 | 1116 |
| LMArena French | 1096 | 1166 |
| LMArena German | 1057 | 1141 |
| LMArena Japanese | 990 | 1037 |
| LMArena Korean | 993 | 1057 |
| LMArena Russian | 1078 | 1158 |
| LMArena Spanish | 1064 | 1151 |
Instruction Following Too close to call
DBRX: 57.5 (#270), Mixtral 8x22B: 57.7 (#266)
| Benchmark | DBRX | Mixtral 8x22B |
|---|---|---|
| LMArena Instruction Following | 1112 | 1147 |
| IFEval | — | 72.4% |
Long Context Too close to call
DBRX: 33.7 (#258), Mixtral 8x22B: 34.7 (#247)
| Benchmark | DBRX | Mixtral 8x22B |
|---|---|---|
| LMArena Longer Query | 1112 | 1144 |
Writing & Preference Mixtral 8x22B leads
DBRX: 33.6 (#275), Mixtral 8x22B: 36.9 (#262)
| Benchmark | DBRX | Mixtral 8x22B |
|---|---|---|
| LMArena Text | 1119 | 1162 |
| LMArena Creative Writing | 1104 | 1141 |
| LMArena Multi-Turn | 1111 | 1130 |
| WildBench | — | 71.1% |
Frequently asked questions
Is DBRX better than Mixtral 8x22B?
DBRX is the stronger model overall, scoring 29.4 to 27.1 on the Noometry Index.
Is DBRX or Mixtral 8x22B better for coding?
DBRX scores higher on coding benchmarks: 32.9 versus 24.2 in the Noometry coding category.
How many benchmarks do DBRX and Mixtral 8x22B share?
21 benchmarks have published results for both models. DBRX has 21 scored results on Noometry and Mixtral 8x22B has 34.