Model comparison
DBRX vs DeepSeek LLM 67B
DBRX is the stronger model overall, scoring 29.4 to 24.9 on the Noometry Index.
Last verified . 12 shared benchmarks.
Databricks
DeepSeekSummary
- They share 12 benchmarks with published results for both. DBRX scores higher in 7 categories and DeepSeek LLM 67B in 1 category; 6 gaps are clear of the uncertainty.
- The widest gap is in math, where DBRX leads 24.3 to 8.7.
- The biggest single-benchmark swing is GPQA Diamond: 32.9% for DBRX and 24.6% for DeepSeek LLM 67B.
Side by side
| DBRX | DeepSeek LLM 67B | |
|---|---|---|
| Provider | Databricks | DeepSeek |
| Noometry Index | 29.4 | 24.9 |
| Released | 2024-03-27 | 2023-11-29 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 21 | 15 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DBRX leads
DBRX: 32.9 (#266), DeepSeek LLM 67B: 31.9 (#278)
| Benchmark | DBRX | DeepSeek LLM 67B |
|---|---|---|
| LMArena Coding | 1132 | 1096 |
| HumanEval+ | 70.1% | — |
| MBPP+ | 55.8% | — |
Reasoning DBRX leads
DBRX: 21.4 (#222), DeepSeek LLM 67B: 16.5 (#304)
| Benchmark | DBRX | DeepSeek LLM 67B |
|---|---|---|
| LMArena Hard Prompts | 1113 | 1070 |
| Chess Puzzles | — | 0% |
| Epoch Capabilities Index | — | 110.5 |
Math DBRX leads
DBRX: 24.3 (#269), DeepSeek LLM 67B: 8.7 (#324)
| Benchmark | DBRX | DeepSeek LLM 67B |
|---|---|---|
| LMArena Math | 1145 | 1108 |
| MATH Level 5 | 11.7% | 6.4% |
| OTIS Mock AIME 2024-2025 | — | 0.8% |
Knowledge DBRX leads
DBRX: 14.9 (#294), DeepSeek LLM 67B: 7.0 (#313)
| Benchmark | DBRX | DeepSeek LLM 67B |
|---|---|---|
| GPQA Diamond | 32.9% | 24.6% |
| LMArena Expert | 1076 | — |
Multilingual Too close to call
DBRX: 29.3 (#268), DeepSeek LLM 67B: 29.4 (#267)
| Benchmark | DBRX | DeepSeek LLM 67B |
|---|---|---|
| LMArena Non-English | 1071 | 1073 |
| LMArena Chinese | 1068 | 1132 |
| 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), DeepSeek LLM 67B: 55.4 (#277)
| Benchmark | DBRX | DeepSeek LLM 67B |
|---|---|---|
| LMArena Instruction Following | 1112 | 1079 |
Long Context Too close to call
DBRX: 33.7 (#258), DeepSeek LLM 67B: 33.1 (#265)
| Benchmark | DBRX | DeepSeek LLM 67B |
|---|---|---|
| LMArena Longer Query | 1112 | 1092 |
Writing & Preference DBRX leads
DBRX: 33.6 (#275), DeepSeek LLM 67B: 31.6 (#282)
| Benchmark | DBRX | DeepSeek LLM 67B |
|---|---|---|
| LMArena Text | 1119 | 1105 |
| LMArena Creative Writing | 1104 | 1067 |
| LMArena Multi-Turn | 1111 | 1082 |
Frequently asked questions
Is DBRX better than DeepSeek LLM 67B?
DBRX is the stronger model overall, scoring 29.4 to 24.9 on the Noometry Index.
Is DBRX or DeepSeek LLM 67B better for coding?
DBRX scores higher on coding benchmarks: 32.9 versus 31.9 in the Noometry coding category.
How many benchmarks do DBRX and DeepSeek LLM 67B share?
12 benchmarks have published results for both models. DBRX has 21 scored results on Noometry and DeepSeek LLM 67B has 15.