Model comparison
Gemma 4 31B IT vs gpt-oss-20b
Gemma 4 31B IT is the stronger model overall, scoring 43.2 to 32.1 on the Noometry Index. gpt-oss-20b costs 4.2× less per token, which makes it the better buy when Gemma 4 31B IT's lead doesn't matter for your workload.
Last verified . 25 shared benchmarks.
GoogleSummary
- They share 25 benchmarks with published results for both. Gemma 4 31B IT scores higher in 8 categories and gpt-oss-20b in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Gemma 4 31B IT leads 60.7 to 35.8.
- The biggest single-benchmark swing is LMCA: 39.3% for Gemma 4 31B IT and 14.5% for gpt-oss-20b.
- gpt-oss-20b is cheaper at $0.018 / $0.09 per million input/output tokens, against $0.09 / $0.34 for Gemma 4 31B IT.
- Gemma 4 31B IT accepts more context: 262K tokens versus 131K.
Side by side
| Gemma 4 31B IT | gpt-oss-20b | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 43.2 | 32.1 |
| Released | 2026-04-02 | 2025-08-05 |
| Weights | Open | Open |
| Context window | 262K | 131K |
| Max output | 33K | 16K |
| Input $ / M tokens | $0.09 | $0.018 |
| Output $ / M tokens | $0.34 | $0.09 |
| Results tracked | 35 | 34 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemma 4 31B IT leads
Gemma 4 31B IT: 41.9 (#98), gpt-oss-20b: 37.1 (#185)
| Benchmark | Gemma 4 31B IT | gpt-oss-20b |
|---|---|---|
| SciCode | 43.4% | 34.4% |
| WeirdML | 52.3% | 40.9% |
| LMArena Coding | 1459 | 1307 |
| ALE-Bench | 925.5 | 566.05 |
| LMArena WebDev | 1364 | — |
Agentic & Tool Use Not comparable
Gemma 4 31B IT: —, gpt-oss-20b: 9.2 (#146)
| Benchmark | Gemma 4 31B IT | gpt-oss-20b |
|---|---|---|
| Terminal-Bench | — | 3.4% |
Reasoning Gemma 4 31B IT leads
Gemma 4 31B IT: 26.4 (#119), gpt-oss-20b: 18.8 (#251)
| Benchmark | Gemma 4 31B IT | gpt-oss-20b |
|---|---|---|
| Kagi LLM Benchmark | 63.5% | 53.2% |
| CritPt | 1.4% | 1.4% |
| Chess Puzzles | 5% | 4% |
| LMArena Hard Prompts | 1448 | 1274 |
| DTBench | 82.7% | 68% |
| LMCA | 39.3% | 14.5% |
| Epoch Capabilities Index | 142.74 | 137.81 |
| NYT Connections (extended) | 70.6% | — |
| Thematic Generalization | 53% | — |
| Surface Evolver Bench | 30.6% | — |
Math Gemma 4 31B IT leads
Gemma 4 31B IT: 41.3 (#72), gpt-oss-20b: 37.6 (#93)
| Benchmark | Gemma 4 31B IT | gpt-oss-20b |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 73.3% | 65.3% |
| LMArena Math | 1464 | 1317 |
| Omni-MATH | — | 56.5% |
Knowledge Gemma 4 31B IT leads
Gemma 4 31B IT: 37.9 (#141), gpt-oss-20b: 34.6 (#186)
| Benchmark | Gemma 4 31B IT | gpt-oss-20b |
|---|---|---|
| GPQA Diamond | 75.8% | 60.8% |
| LMArena Expert | 1465 | 1258 |
| SimpleQA Verified | 10.4% | — |
| MMLU-Pro | — | 74% |
| Vectara Hallucination Rate | 7.4% | — |
| GPQA (HELM) | — | 59.4% |
Multimodal Not comparable
Gemma 4 31B IT: 41.0 (#32), gpt-oss-20b: —
| Benchmark | Gemma 4 31B IT | gpt-oss-20b |
|---|---|---|
| LMArena Vision | 1278 | — |
| LMArena Document | 1425 | — |
Multilingual Gemma 4 31B IT leads
Gemma 4 31B IT: 54.0 (#52), gpt-oss-20b: 42.4 (#192)
| Benchmark | Gemma 4 31B IT | gpt-oss-20b |
|---|---|---|
| LMArena Non-English | 1431 | 1269 |
| LMArena Chinese | 1478 | 1311 |
| LMArena Russian | 1462 | 1279 |
| LMArena Spanish | 1443 | 1267 |
| LMArena French | 1436 | — |
| LMArena German | — | 1249 |
| LMArena Japanese | — | 1245 |
| LMArena Korean | — | 1239 |
Instruction Following Gemma 4 31B IT leads
Gemma 4 31B IT: 75.6 (#58), gpt-oss-20b: 62.0 (#235)
| Benchmark | Gemma 4 31B IT | gpt-oss-20b |
|---|---|---|
| LMArena Instruction Following | 1433 | 1236 |
| IFEval | — | 73.2% |
Long Context Gemma 4 31B IT leads
Gemma 4 31B IT: 44.2 (#67), gpt-oss-20b: 37.9 (#203)
| Benchmark | Gemma 4 31B IT | gpt-oss-20b |
|---|---|---|
| LMArena Longer Query | 1446 | 1251 |
Writing & Preference Gemma 4 31B IT leads
Gemma 4 31B IT: 60.7 (#92), gpt-oss-20b: 35.8 (#260)
| Benchmark | Gemma 4 31B IT | gpt-oss-20b |
|---|---|---|
| LMArena Text | 1443 | 1288 |
| LMArena Creative Writing | 1416 | 1201 |
| EQ-Bench Creative Writing | 1368 | 666 |
| LMArena Multi-Turn | 1452 | 1268 |
| WildBench | — | 73.7% |
| EQ-Bench 4 | 1120 | — |
Frequently asked questions
Is Gemma 4 31B IT better than gpt-oss-20b?
Gemma 4 31B IT is the stronger model overall, scoring 43.2 to 32.1 on the Noometry Index. gpt-oss-20b costs 4.2× less per token, which makes it the better buy when Gemma 4 31B IT's lead doesn't matter for your workload.
Which is cheaper, Gemma 4 31B IT or gpt-oss-20b?
gpt-oss-20b is cheaper. It lists at $0.018 per million input tokens and $0.09 per million output tokens; Gemma 4 31B IT lists at $0.09 and $0.34.
Is Gemma 4 31B IT or gpt-oss-20b better for coding?
Gemma 4 31B IT scores higher on coding benchmarks: 41.9 versus 37.1 in the Noometry coding category.
Which has the bigger context window?
Gemma 4 31B IT does, with 262K tokens against 131K.
How many benchmarks do Gemma 4 31B IT and gpt-oss-20b share?
25 benchmarks have published results for both models. Gemma 4 31B IT has 35 scored results on Noometry and gpt-oss-20b has 34.