Model comparison
DeepSeek-V3.2-Exp vs Gemma 2 9B
DeepSeek-V3.2-Exp is the stronger model overall, scoring 44.3 to 25.9 on the Noometry Index.
Last verified . 21 shared benchmarks.
DeepSeek
GoogleSummary
- They share 21 benchmarks with published results for both. DeepSeek-V3.2-Exp scores higher in 8 categories and Gemma 2 9B in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where DeepSeek-V3.2-Exp leads 51.7 to 9.7.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 87.8% for DeepSeek-V3.2-Exp and 0.6% for Gemma 2 9B.
Side by side
| DeepSeek-V3.2-Exp | Gemma 2 9B | |
|---|---|---|
| Provider | DeepSeek | |
| Noometry Index | 44.3 | 25.9 |
| Released | 2025-09-29 | 2024-06-24 |
| Weights | Open | Open |
| Context window | 164K | — |
| Max output | 66K | — |
| Input $ / M tokens | $0.26 | — |
| Output $ / M tokens | $0.38 | — |
| Results tracked | 49 | 35 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 46.5 (#65), Gemma 2 9B: 29.4 (#304)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 2 9B |
|---|---|---|
| LMArena Coding | 1454 | 1173 |
| SWE-bench Verified (bash only) | 70% | — |
| Aider Polyglot | 74.2% | — |
| LMArena WebDev | 1362 | — |
| SWE-bench Multilingual | 59% | — |
| SciCode | 38.9% | — |
| WeirdML | 39.5% | — |
| BigCodeBench Instruct | — | 34.7% |
| LiveBench Coding | — | 22.5% |
| BigCodeBench Complete | — | 40.6% |
Agentic & Tool Use Not comparable
DeepSeek-V3.2-Exp: 32.7 (#59), Gemma 2 9B: —
| Benchmark | DeepSeek-V3.2-Exp | Gemma 2 9B |
|---|---|---|
| Terminal-Bench | 39.6% | — |
| APEX-Agents | 21.3% | — |
| Berkeley Function Calling Leaderboard | 56.7% | — |
| TheAgentCompany | 42.9% | — |
| Vending-Bench 2 | 1,034 | — |
Reasoning DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 22.1 (#208), Gemma 2 9B: 15.9 (#309)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 2 9B |
|---|---|---|
| LMArena Hard Prompts | 1434 | 1171 |
| Epoch Capabilities Index | 146.27 | 119.83 |
| ARC-AGI-2 | 4% | — |
| Kagi LLM Benchmark | 52.2% | — |
| NYT Connections (extended) | 36.7% | — |
| ARC-AGI-1 | 57% | — |
| CritPt | 2.9% | — |
| Chess Puzzles | 14% | — |
| Thematic Generalization | 65% | — |
| LiveBench Reasoning | — | 15.2% |
| DTBench | 87.7% | — |
| LiveBench Data Analysis | — | 36.4% |
| LMCA | 29.1% | — |
| LiveBench | — | 28.7% |
| PIQA | — | 83.7% |
Math DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 41.7 (#87), Gemma 2 9B: 9.9 (#318)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 2 9B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 87.8% | 0.6% |
| LMArena Math | 1435 | 1183 |
| MathArena Final-Answer Competitions | 57.7% | — |
| ProofBench | 8% | — |
| LiveBench Math | — | 19.8% |
| MATH Level 5 | — | 21% |
| FrontierMath (Feb 2025 set) | 22.1% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
| GSM8K | — | 84.9% |
Knowledge DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 51.7 (#66), Gemma 2 9B: 9.7 (#305)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 2 9B |
|---|---|---|
| GPQA Diamond | 83.4% | 27.5% |
| LMArena Expert | 1436 | 1147 |
| Vectara Hallucination Rate | 5.3% | — |
| BoolQ | — | 85.7% |
| MMLU | — | 72.1% |
Multilingual DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 52.2 (#90), Gemma 2 9B: 36.6 (#238)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 2 9B |
|---|---|---|
| LMArena Non-English | 1409 | 1188 |
| LMArena Chinese | 1461 | 1185 |
| LMArena French | 1433 | 1190 |
| LMArena German | 1440 | 1186 |
| LMArena Japanese | 1374 | 1144 |
| LMArena Korean | 1371 | 1137 |
| LMArena Russian | 1424 | 1200 |
| LMArena Spanish | 1440 | 1200 |
Instruction Following DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 74.5 (#93), Gemma 2 9B: 57.6 (#269)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 2 9B |
|---|---|---|
| LMArena Instruction Following | 1413 | 1178 |
| LiveBench Instruction Following | — | 52.6% |
Long Context DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 47.6 (#16), Gemma 2 9B: 36.3 (#233)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 2 9B |
|---|---|---|
| LMArena Longer Query | 1428 | 1197 |
| Fiction.LiveBench | 83.3% | — |
| CL-bench | 13.2% | — |
| CL-bench Life | 9.5% | — |
Writing & Preference DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 62.4 (#77), Gemma 2 9B: 32.1 (#281)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 2 9B |
|---|---|---|
| LMArena Text | 1425 | 1207 |
| LMArena Creative Writing | 1403 | 1206 |
| EQ-Bench Creative Writing | 1515 | 841 |
| LMArena Multi-Turn | 1427 | 1193 |
| LiveBench Language | — | 25.5% |
Frequently asked questions
Is DeepSeek-V3.2-Exp better than Gemma 2 9B?
DeepSeek-V3.2-Exp is the stronger model overall, scoring 44.3 to 25.9 on the Noometry Index.
Is DeepSeek-V3.2-Exp or Gemma 2 9B better for coding?
DeepSeek-V3.2-Exp scores higher on coding benchmarks: 46.5 versus 29.4 in the Noometry coding category.
How many benchmarks do DeepSeek-V3.2-Exp and Gemma 2 9B share?
21 benchmarks have published results for both models. DeepSeek-V3.2-Exp has 49 scored results on Noometry and Gemma 2 9B has 35.