Model comparison
DBRX vs Qwen-14B
Qwen-14B is the stronger model overall, scoring 31.4 to 29.4 on the Noometry Index.
Last verified . 10 shared benchmarks.
Databricks
Alibaba (Qwen)Summary
- They share 10 benchmarks with published results for both. DBRX scores higher in 6 categories and Qwen-14B in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen-14B leads 31.2 to 24.3.
Side by side
| DBRX | Qwen-14B | |
|---|---|---|
| Provider | Databricks | Alibaba (Qwen) |
| Noometry Index | 29.4 | 31.4 |
| Released | 2024-03-27 | 2023-09-24 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 21 | 18 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DBRX leads
DBRX: 32.9 (#266), Qwen-14B: 31.2 (#288)
| Benchmark | DBRX | Qwen-14B |
|---|---|---|
| LMArena Coding | 1132 | 1071 |
| HumanEval+ | 70.1% | — |
| MBPP+ | 55.8% | — |
Reasoning DBRX leads
DBRX: 21.4 (#222), Qwen-14B: 19.6 (#257)
| Benchmark | DBRX | Qwen-14B |
|---|---|---|
| LMArena Hard Prompts | 1113 | 1027 |
| BIG-Bench Hard | — | 55% |
| Epoch Capabilities Index | — | 113.03 |
| LAMBADA | — | 71.1% |
| PIQA | — | 79.9% |
Math Qwen-14B leads
DBRX: 24.3 (#269), Qwen-14B: 31.2 (#227)
| Benchmark | DBRX | Qwen-14B |
|---|---|---|
| LMArena Math | 1145 | 1068 |
| MATH Level 5 | 11.7% | — |
| GSM8K | — | 61.3% |
Knowledge Not comparable
DBRX: 14.9 (#294), Qwen-14B: —
| Benchmark | DBRX | Qwen-14B |
|---|---|---|
| GPQA Diamond | 32.9% | — |
| LMArena Expert | 1076 | — |
| ARC (AI2) Challenge | — | 84.4% |
| BoolQ | — | 86.2% |
| MMLU | — | 66.3% |
Multilingual DBRX leads
DBRX: 29.3 (#268), Qwen-14B: 27.5 (#275)
| Benchmark | DBRX | Qwen-14B |
|---|---|---|
| LMArena Non-English | 1071 | 1041 |
| LMArena Chinese | 1068 | 1077 |
| LMArena French | 1096 | — |
| LMArena German | 1057 | — |
| LMArena Japanese | 990 | — |
| LMArena Korean | 993 | — |
| LMArena Russian | 1078 | — |
| LMArena Spanish | 1064 | — |
Instruction Following DBRX leads
DBRX: 57.5 (#270), Qwen-14B: 52.4 (#289)
| Benchmark | DBRX | Qwen-14B |
|---|---|---|
| LMArena Instruction Following | 1112 | 1031 |
Long Context DBRX leads
DBRX: 33.7 (#258), Qwen-14B: 31.3 (#280)
| Benchmark | DBRX | Qwen-14B |
|---|---|---|
| LMArena Longer Query | 1112 | 1028 |
Writing & Preference DBRX leads
DBRX: 33.6 (#275), Qwen-14B: 27.6 (#299)
| Benchmark | DBRX | Qwen-14B |
|---|---|---|
| LMArena Text | 1119 | 1051 |
| LMArena Creative Writing | 1104 | 1028 |
| LMArena Multi-Turn | 1111 | 1022 |
Frequently asked questions
Is DBRX better than Qwen-14B?
Qwen-14B is the stronger model overall, scoring 31.4 to 29.4 on the Noometry Index.
Is DBRX or Qwen-14B better for coding?
DBRX scores higher on coding benchmarks: 32.9 versus 31.2 in the Noometry coding category.
How many benchmarks do DBRX and Qwen-14B share?
10 benchmarks have published results for both models. DBRX has 21 scored results on Noometry and Qwen-14B has 18.