Model comparison
DBRX vs DeepSeek V4 Pro
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 29.4 on the Noometry Index.
Last verified . 18 shared benchmarks.
Databricks
DeepSeekSummary
- They share 18 benchmarks with published results for both. DBRX scores higher in 0 categories and DeepSeek V4 Pro in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where DeepSeek V4 Pro leads 59.5 to 14.9.
- The biggest single-benchmark swing is GPQA Diamond: 32.9% for DBRX and 91.7% for DeepSeek V4 Pro.
Side by side
| DBRX | DeepSeek V4 Pro | |
|---|---|---|
| Provider | Databricks | DeepSeek |
| Noometry Index | 29.4 | 54.3 |
| Released | 2024-03-27 | 2026-04-24 |
| Weights | Open | Open |
| Context window | — | 1M |
| Max output | — | 393K |
| Input $ / M tokens | — | $0.66 |
| Output $ / M tokens | — | $1.98 |
| Results tracked | 21 | 48 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek V4 Pro leads
DBRX: 32.9 (#266), DeepSeek V4 Pro: 52.4 (#34)
| Benchmark | DBRX | DeepSeek V4 Pro |
|---|---|---|
| LMArena Coding | 1132 | 1470 |
| SWE-bench Verified | — | 77.6% |
| FrontierCode | — | 28.6% |
| LMArena WebDev | — | 1582 |
| SciCode | — | 51% |
| WeirdML | — | 66.2% |
| ALE-Bench | — | 1,403 |
| HumanEval+ | 70.1% | — |
| MBPP+ | 55.8% | — |
Agentic & Tool Use Not comparable
DBRX: —, DeepSeek V4 Pro: 32.8 (#58)
| Benchmark | DBRX | DeepSeek V4 Pro |
|---|---|---|
| APEX-Agents | — | 47.3% |
| Vending-Bench 2 | — | 3,285 |
Reasoning DeepSeek V4 Pro leads
DBRX: 21.4 (#222), DeepSeek V4 Pro: 56.5 (#24)
| Benchmark | DBRX | DeepSeek V4 Pro |
|---|---|---|
| LMArena Hard Prompts | 1113 | 1461 |
| ARC-AGI-2 | — | 61.3% |
| Kagi LLM Benchmark | — | 53.5% |
| NYT Connections (extended) | — | 91.3% |
| ARC-AGI-1 | — | 90.5% |
| CritPt | — | 18% |
| Chess Puzzles | — | 47% |
| Mystery Game Puzzles | — | 43% |
| DTBench | — | 93.9% |
| LMCA | — | 45.5% |
| Surface Evolver Bench | — | 40% |
| Epoch Capabilities Index | — | 155.31 |
| ForecastBench | — | 56.1 |
Math DeepSeek V4 Pro leads
DBRX: 24.3 (#269), DeepSeek V4 Pro: 64.8 (#30)
| Benchmark | DBRX | DeepSeek V4 Pro |
|---|---|---|
| LMArena Math | 1145 | 1455 |
| FrontierMath (Tiers 1-3) | — | 64.6% |
| FrontierMath Tier 4 | — | 26.8% |
| MathArena Final-Answer Competitions | — | 76.6% |
| OTIS Mock AIME 2024-2025 | — | 98.6% |
| ProofBench | — | 50% |
| MATH Level 5 | 11.7% | — |
Knowledge DeepSeek V4 Pro leads
DBRX: 14.9 (#294), DeepSeek V4 Pro: 59.5 (#31)
| Benchmark | DBRX | DeepSeek V4 Pro |
|---|---|---|
| GPQA Diamond | 32.9% | 91.7% |
| LMArena Expert | 1076 | 1464 |
| SimpleQA Verified | — | 52.9% |
| Vectara Hallucination Rate | — | 8.6% |
Multilingual DeepSeek V4 Pro leads
DBRX: 29.3 (#268), DeepSeek V4 Pro: 54.4 (#45)
| Benchmark | DBRX | DeepSeek V4 Pro |
|---|---|---|
| LMArena Non-English | 1071 | 1439 |
| LMArena Chinese | 1068 | 1486 |
| LMArena French | 1096 | 1472 |
| LMArena German | 1057 | 1458 |
| LMArena Japanese | 990 | 1445 |
| LMArena Korean | 993 | 1447 |
| LMArena Russian | 1078 | 1453 |
| LMArena Spanish | 1064 | 1458 |
Instruction Following DeepSeek V4 Pro leads
DBRX: 57.5 (#270), DeepSeek V4 Pro: 76.1 (#47)
| Benchmark | DBRX | DeepSeek V4 Pro |
|---|---|---|
| LMArena Instruction Following | 1112 | 1448 |
Long Context DeepSeek V4 Pro leads
DBRX: 33.7 (#258), DeepSeek V4 Pro: 45.0 (#51)
| Benchmark | DBRX | DeepSeek V4 Pro |
|---|---|---|
| LMArena Longer Query | 1112 | 1458 |
| CL-bench Life | — | 13.5% |
Writing & Preference DeepSeek V4 Pro leads
DBRX: 33.6 (#275), DeepSeek V4 Pro: 65.5 (#46)
| Benchmark | DBRX | DeepSeek V4 Pro |
|---|---|---|
| LMArena Text | 1119 | 1451 |
| LMArena Creative Writing | 1104 | 1446 |
| LMArena Multi-Turn | 1111 | 1467 |
| EQ-Bench Creative Writing | — | 1553 |
| EQ-Bench 4 | — | 1166 |
Frequently asked questions
Is DBRX better than DeepSeek V4 Pro?
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 29.4 on the Noometry Index.
Is DBRX or DeepSeek V4 Pro better for coding?
DeepSeek V4 Pro scores higher on coding benchmarks: 52.4 versus 32.9 in the Noometry coding category.
How many benchmarks do DBRX and DeepSeek V4 Pro share?
18 benchmarks have published results for both models. DBRX has 21 scored results on Noometry and DeepSeek V4 Pro has 48.