Model comparison
DeepSeek V4 Pro vs Gemma 2B
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 29.6 on the Noometry Index.
Last verified . 12 shared benchmarks.
DeepSeek
GoogleSummary
- They share 12 benchmarks with published results for both. DeepSeek V4 Pro scores higher in 7 categories and Gemma 2B in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where DeepSeek V4 Pro leads 65.5 to 24.0.
Side by side
| DeepSeek V4 Pro | Gemma 2B | |
|---|---|---|
| Provider | DeepSeek | |
| Noometry Index | 54.3 | 29.6 |
| Released | 2026-04-24 | 2024-02-21 |
| Weights | Open | Open |
| Context window | 1M | — |
| Max output | 393K | — |
| Input $ / M tokens | $0.66 | — |
| Output $ / M tokens | $1.98 | — |
| Results tracked | 48 | 23 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek V4 Pro leads
DeepSeek V4 Pro: 52.4 (#34), Gemma 2B: 29.4 (#305)
| Benchmark | DeepSeek V4 Pro | Gemma 2B |
|---|---|---|
| LMArena Coding | 1470 | 1010 |
| SWE-bench Verified | 77.6% | — |
| FrontierCode | 28.6% | — |
| LMArena WebDev | 1582 | — |
| SciCode | 51% | — |
| WeirdML | 66.2% | — |
| ALE-Bench | 1,403 | — |
| HumanEval+ | — | 20.7% |
| MBPP+ | — | 34.1% |
Agentic & Tool Use Not comparable
DeepSeek V4 Pro: 32.8 (#58), Gemma 2B: —
| Benchmark | DeepSeek V4 Pro | Gemma 2B |
|---|---|---|
| APEX-Agents | 47.3% | — |
| Vending-Bench 2 | 3,285 | — |
Reasoning DeepSeek V4 Pro leads
DeepSeek V4 Pro: 56.5 (#24), Gemma 2B: 18.8 (#275)
| Benchmark | DeepSeek V4 Pro | Gemma 2B |
|---|---|---|
| LMArena Hard Prompts | 1461 | 989 |
| Epoch Capabilities Index | 155.31 | 94.2 |
| 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% | — |
| BIG-Bench Hard | — | 35.2% |
| ForecastBench | 56.1 | — |
| HellaSwag | — | 71.4% |
| PIQA | — | 77.3% |
| WinoGrande | — | 65.4% |
Math DeepSeek V4 Pro leads
DeepSeek V4 Pro: 64.8 (#30), Gemma 2B: 30.0 (#239)
| Benchmark | DeepSeek V4 Pro | Gemma 2B |
|---|---|---|
| LMArena Math | 1455 | 1009 |
| 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% | — |
| GSM8K | — | 17.7% |
Knowledge Not comparable
DeepSeek V4 Pro: 59.5 (#31), Gemma 2B: —
| Benchmark | DeepSeek V4 Pro | Gemma 2B |
|---|---|---|
| GPQA Diamond | 91.7% | — |
| SimpleQA Verified | 52.9% | — |
| Vectara Hallucination Rate | 8.6% | — |
| LMArena Expert | 1464 | — |
| ARC (AI2) Challenge | — | 42.1% |
| BoolQ | — | 69.4% |
| MMLU | — | 42.3% |
| TriviaQA | — | 53.2% |
Multilingual DeepSeek V4 Pro leads
DeepSeek V4 Pro: 54.4 (#45), Gemma 2B: 23.0 (#294)
| Benchmark | DeepSeek V4 Pro | Gemma 2B |
|---|---|---|
| LMArena Non-English | 1439 | 958 |
| LMArena Chinese | 1486 | 986 |
| LMArena Russian | 1453 | 937 |
| LMArena French | 1472 | — |
| LMArena German | 1458 | — |
| LMArena Japanese | 1445 | — |
| LMArena Korean | 1447 | — |
| LMArena Spanish | 1458 | — |
Instruction Following DeepSeek V4 Pro leads
DeepSeek V4 Pro: 76.1 (#47), Gemma 2B: 48.5 (#302)
| Benchmark | DeepSeek V4 Pro | Gemma 2B |
|---|---|---|
| LMArena Instruction Following | 1448 | 970 |
Long Context DeepSeek V4 Pro leads
DeepSeek V4 Pro: 45.0 (#51), Gemma 2B: 29.9 (#291)
| Benchmark | DeepSeek V4 Pro | Gemma 2B |
|---|---|---|
| LMArena Longer Query | 1458 | 981 |
| CL-bench Life | 13.5% | — |
Writing & Preference DeepSeek V4 Pro leads
DeepSeek V4 Pro: 65.5 (#46), Gemma 2B: 24.0 (#308)
| Benchmark | DeepSeek V4 Pro | Gemma 2B |
|---|---|---|
| LMArena Text | 1451 | 1002 |
| LMArena Creative Writing | 1446 | 987 |
| LMArena Multi-Turn | 1467 | 945 |
| EQ-Bench Creative Writing | 1553 | — |
| EQ-Bench 4 | 1166 | — |
Frequently asked questions
Is DeepSeek V4 Pro better than Gemma 2B?
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 29.6 on the Noometry Index.
Is DeepSeek V4 Pro or Gemma 2B better for coding?
DeepSeek V4 Pro scores higher on coding benchmarks: 52.4 versus 29.4 in the Noometry coding category.
How many benchmarks do DeepSeek V4 Pro and Gemma 2B share?
12 benchmarks have published results for both models. DeepSeek V4 Pro has 48 scored results on Noometry and Gemma 2B has 23.