Model comparison
DeepSeek-V3.2-Exp vs Gemma 4 31B IT
DeepSeek-V3.2-Exp and Gemma 4 31B IT score almost the same on the Noometry Index (44.3 vs 43.5), so choose on price, context window or the category you care about most.
Last verified . 29 shared benchmarks.
DeepSeek
GoogleSummary
- They share 29 benchmarks with published results for both. DeepSeek-V3.2-Exp scores higher in 4 categories and Gemma 4 31B IT in 4 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where DeepSeek-V3.2-Exp leads 51.7 to 37.9.
- The biggest single-benchmark swing is NYT Connections (extended): 36.7% for DeepSeek-V3.2-Exp and 70.6% for Gemma 4 31B IT.
- Gemma 4 31B IT is cheaper at $0.09 / $0.34 per million input/output tokens, against $0.26 / $0.38 for DeepSeek-V3.2-Exp.
- Gemma 4 31B IT accepts more context: 262K tokens versus 164K.
Side by side
| DeepSeek-V3.2-Exp | Gemma 4 31B IT | |
|---|---|---|
| Provider | DeepSeek | |
| Noometry Index | 44.3 | 43.5 |
| Released | 2025-09-29 | 2026-04-02 |
| Weights | Open | Open |
| Context window | 164K | 262K |
| Max output | 66K | 33K |
| Input $ / M tokens | $0.26 | $0.09 |
| Output $ / M tokens | $0.38 | $0.34 |
| 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 4 31B IT: 42.3 (#108)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 4 31B IT |
|---|---|---|
| LMArena WebDev | 1362 | 1366 |
| SciCode | 38.9% | 43.4% |
| WeirdML | 39.5% | 52.3% |
| LMArena Coding | 1454 | 1459 |
| SWE-bench Verified (bash only) | 70% | — |
| Aider Polyglot | 74.2% | — |
| SWE-bench Multilingual | 59% | — |
| ALE-Bench | — | 925.5 |
Agentic & Tool Use Not comparable
DeepSeek-V3.2-Exp: 32.7 (#59), Gemma 4 31B IT: —
| Benchmark | DeepSeek-V3.2-Exp | Gemma 4 31B IT |
|---|---|---|
| Terminal-Bench | 39.6% | — |
| APEX-Agents | 21.3% | — |
| Berkeley Function Calling Leaderboard | 56.7% | — |
| TheAgentCompany | 42.9% | — |
| Vending-Bench 2 | 1,034 | — |
Reasoning Gemma 4 31B IT leads
DeepSeek-V3.2-Exp: 22.1 (#208), Gemma 4 31B IT: 27.2 (#122)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 4 31B IT |
|---|---|---|
| Kagi LLM Benchmark | 52.2% | 63.5% |
| NYT Connections (extended) | 36.7% | 70.6% |
| CritPt | 2.9% | 1.4% |
| Chess Puzzles | 14% | 5% |
| Thematic Generalization | 65% | 53% |
| LMArena Hard Prompts | 1434 | 1448 |
| DTBench | 87.7% | 82.7% |
| LMCA | 29.1% | 39.3% |
| Epoch Capabilities Index | 146.27 | 142.74 |
| ARC-AGI-2 | 4% | — |
| ARC-AGI-1 | 57% | — |
| Surface Evolver Bench | — | 30.6% |
Math Gemma 4 31B IT leads
DeepSeek-V3.2-Exp: 41.7 (#87), Gemma 4 31B IT: 43.2 (#81)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 4 31B IT |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 87.8% | 73.3% |
| LMArena Math | 1435 | 1465 |
| 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 4 31B IT: 37.9 (#151)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 4 31B IT |
|---|---|---|
| GPQA Diamond | 83.4% | 75.8% |
| Vectara Hallucination Rate | 5.3% | 7.4% |
| LMArena Expert | 1436 | 1465 |
| SimpleQA Verified | — | 10.4% |
Multimodal Not comparable
DeepSeek-V3.2-Exp: —, Gemma 4 31B IT: 41.6 (#34)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 4 31B IT |
|---|---|---|
| LMArena Vision | — | 1277 |
| LMArena Document | — | 1425 |
Multilingual Gemma 4 31B IT leads
DeepSeek-V3.2-Exp: 52.2 (#90), Gemma 4 31B IT: 53.8 (#57)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 4 31B IT |
|---|---|---|
| LMArena Non-English | 1409 | 1431 |
| LMArena Chinese | 1461 | 1476 |
| LMArena French | 1433 | 1435 |
| LMArena Russian | 1424 | 1460 |
| LMArena Spanish | 1440 | 1444 |
| LMArena German | 1440 | — |
| LMArena Japanese | 1374 | — |
| LMArena Korean | 1371 | — |
Instruction Following Too close to call
DeepSeek-V3.2-Exp: 74.5 (#93), Gemma 4 31B IT: 75.5 (#61)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 4 31B IT |
|---|---|---|
| LMArena Instruction Following | 1413 | 1433 |
Long Context DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 47.6 (#16), Gemma 4 31B IT: 44.2 (#71)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 4 31B IT |
|---|---|---|
| LMArena Longer Query | 1428 | 1446 |
| 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 4 31B IT: 60.5 (#96)
| Benchmark | DeepSeek-V3.2-Exp | Gemma 4 31B IT |
|---|---|---|
| LMArena Text | 1425 | 1443 |
| LMArena Creative Writing | 1403 | 1415 |
| EQ-Bench Creative Writing | 1515 | 1368 |
| LMArena Multi-Turn | 1427 | 1452 |
| EQ-Bench 4 | — | 1120 |
Frequently asked questions
Is DeepSeek-V3.2-Exp better than Gemma 4 31B IT?
DeepSeek-V3.2-Exp and Gemma 4 31B IT score almost the same on the Noometry Index (44.3 vs 43.5), so choose on price, context window or the category you care about most.
Which is cheaper, DeepSeek-V3.2-Exp or Gemma 4 31B IT?
Gemma 4 31B IT is cheaper. It lists at $0.09 per million input tokens and $0.34 per million output tokens; DeepSeek-V3.2-Exp lists at $0.26 and $0.38.
Is DeepSeek-V3.2-Exp or Gemma 4 31B IT better for coding?
DeepSeek-V3.2-Exp scores higher on coding benchmarks: 46.5 versus 42.3 in the Noometry coding category.
Which has the bigger context window?
Gemma 4 31B IT does, with 262K tokens against 164K.
How many benchmarks do DeepSeek-V3.2-Exp and Gemma 4 31B IT share?
29 benchmarks have published results for both models. DeepSeek-V3.2-Exp has 49 scored results on Noometry and Gemma 4 31B IT has 35.