Model comparison
DeepSeek-V3.2-Exp vs Gemma 3 4B
DeepSeek-V3.2-Exp is the stronger model overall, scoring 44.3 to 28.1 on the Noometry Index. Gemma 3 4B costs 5.8× less per token, which makes it the better buy when DeepSeek-V3.2-Exp's lead doesn't matter for your workload.
Last verified . 22 shared benchmarks.
DeepSeek
GoogleSummary
- They share 22 benchmarks with published results for both. DeepSeek-V3.2-Exp scores higher in 9 categories and Gemma 3 4B in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where DeepSeek-V3.2-Exp leads 51.7 to 11.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 87.8% for DeepSeek-V3.2-Exp and 7.5% for Gemma 3 4B.
- Gemma 3 4B is cheaper at $0.04 / $0.08 per million input/output tokens, against $0.26 / $0.38 for DeepSeek-V3.2-Exp.
- DeepSeek-V3.2-Exp accepts more context: 164K tokens versus 131K.
Side by side
| DeepSeek-V3.2-Exp | Gemma 3 4B | |
|---|---|---|
| Provider | DeepSeek | |
| Noometry Index | 44.3 | 28.1 |
| Released | 2025-09-29 | 2025-03-12 |
| Weights | Open | Open |
| Context window | 164K | 131K |
| Max output | 66K | 4K |
| Input $ / M tokens | $0.26 | $0.04 |
| Output $ / M tokens | $0.38 | $0.08 |
| Results tracked | 49 | 22 |
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 3 4B: 35.9 (#215)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 3 4B |
|---|---|---|
| LMArena Coding | 1454 | 1230 |
| SWE-bench Verified (bash only) | 70% | — |
| Aider Polyglot | 74.2% | — |
| LMArena WebDev | 1362 | — |
| SWE-bench Multilingual | 59% | — |
| SciCode | 38.9% | — |
| WeirdML | 39.5% | — |
Agentic & Tool Use DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 32.7 (#59), Gemma 3 4B: 20.9 (#142)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 3 4B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 56.7% | 19.6% |
| Terminal-Bench | 39.6% | — |
| APEX-Agents | 21.3% | — |
| TheAgentCompany | 42.9% | — |
| Vending-Bench 2 | 1,034 | — |
Reasoning DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 22.1 (#208), Gemma 3 4B: 13.2 (#335)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 3 4B |
|---|---|---|
| Kagi LLM Benchmark | 52.2% | 25.2% |
| Chess Puzzles | 14% | 0% |
| LMArena Hard Prompts | 1434 | 1253 |
| DTBench | 87.7% | 50.9% |
| LMCA | 29.1% | 2.8% |
| Epoch Capabilities Index | 146.27 | 116.02 |
| ARC-AGI-2 | 4% | — |
| NYT Connections (extended) | 36.7% | — |
| ARC-AGI-1 | 57% | — |
| CritPt | 2.9% | — |
| Thematic Generalization | 65% | — |
Math DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 41.7 (#87), Gemma 3 4B: 16.8 (#292)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 3 4B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 87.8% | 7.5% |
| LMArena Math | 1435 | 1239 |
| MathArena Final-Answer Competitions | 57.7% | — |
| ProofBench | 8% | — |
| FrontierMath (Feb 2025 set) | 22.1% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 51.7 (#66), Gemma 3 4B: 11.8 (#299)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 3 4B |
|---|---|---|
| GPQA Diamond | 83.4% | 23.2% |
| Vectara Hallucination Rate | 5.3% | 6.4% |
| LMArena Expert | 1436 | 1223 |
Multilingual DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 52.2 (#90), Gemma 3 4B: 42.5 (#194)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 3 4B |
|---|---|---|
| LMArena Non-English | 1409 | 1273 |
| LMArena German | 1440 | 1281 |
| LMArena Russian | 1424 | 1294 |
| LMArena Chinese | 1461 | — |
| LMArena French | 1433 | — |
| LMArena Japanese | 1374 | — |
| LMArena Korean | 1371 | — |
| LMArena Spanish | 1440 | — |
Instruction Following DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 74.5 (#93), Gemma 3 4B: 65.2 (#225)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 3 4B |
|---|---|---|
| LMArena Instruction Following | 1413 | 1239 |
Long Context DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 47.6 (#16), Gemma 3 4B: 38.7 (#194)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 3 4B |
|---|---|---|
| LMArena Longer Query | 1428 | 1273 |
| 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 3 4B: 42.0 (#239)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 3 4B |
|---|---|---|
| LMArena Text | 1425 | 1291 |
| LMArena Creative Writing | 1403 | 1271 |
| EQ-Bench Creative Writing | 1515 | 1068 |
| LMArena Multi-Turn | 1427 | 1255 |
Frequently asked questions
Is DeepSeek-V3.2-Exp better than Gemma 3 4B?
DeepSeek-V3.2-Exp is the stronger model overall, scoring 44.3 to 28.1 on the Noometry Index. Gemma 3 4B costs 5.8× less per token, which makes it the better buy when DeepSeek-V3.2-Exp's lead doesn't matter for your workload.
Which is cheaper, DeepSeek-V3.2-Exp or Gemma 3 4B?
Gemma 3 4B is cheaper. It lists at $0.04 per million input tokens and $0.08 per million output tokens; DeepSeek-V3.2-Exp lists at $0.26 and $0.38.
Is DeepSeek-V3.2-Exp or Gemma 3 4B better for coding?
DeepSeek-V3.2-Exp scores higher on coding benchmarks: 46.5 versus 35.9 in the Noometry coding category.
Which has the bigger context window?
DeepSeek-V3.2-Exp does, with 164K tokens against 131K.
How many benchmarks do DeepSeek-V3.2-Exp and Gemma 3 4B share?
22 benchmarks have published results for both models. DeepSeek-V3.2-Exp has 49 scored results on Noometry and Gemma 3 4B has 22.