Model comparison
Gemma 4 31B IT vs GPT-5
GPT-5 is the stronger model overall, scoring 50.9 to 43.5 on the Noometry Index. Gemma 4 31B IT costs 23× less per token, which makes it the better buy when GPT-5's lead doesn't matter for your workload.
Last verified . 30 shared benchmarks.
GoogleSummary
- They share 30 benchmarks with published results for both. Gemma 4 31B IT scores higher in 2 categories and GPT-5 in 7 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in long context, where GPT-5 leads 69.5 to 44.2.
- The biggest single-benchmark swing is SimpleQA Verified: 10.4% for Gemma 4 31B IT and 50.1% for GPT-5.
- Gemma 4 31B IT is cheaper at $0.09 / $0.34 per million input/output tokens, against $1.25 / $10 for GPT-5.
- GPT-5 accepts more context: 400K tokens versus 262K.
- Gemma 4 31B IT has downloadable open weights; the other is API-only.
Side by side
| Gemma 4 31B IT | GPT-5 | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 43.5 | 50.9 |
| Released | 2026-04-02 | 2025-08-07 |
| Weights | Open | Proprietary |
| Context window | 262K | 400K |
| Max output | 33K | 128K |
| Input $ / M tokens | $0.09 | $1.25 |
| Output $ / M tokens | $0.34 | $10 |
| Results tracked | 35 | 69 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5 leads
Gemma 4 31B IT: 42.3 (#108), GPT-5: 50.3 (#47)
| Benchmark | Gemma 4 31B IT | GPT-5 |
|---|---|---|
| LMArena WebDev | 1366 | 1418 |
| SciCode | 43.4% | 42.9% |
| WeirdML | 52.3% | 60.7% |
| LMArena Coding | 1459 | 1436 |
| ALE-Bench | 925.5 | 1,162 |
| SWE-bench Verified | — | 73.6% |
| SWE-bench Verified (bash only) | — | 65% |
| Aider Polyglot | — | 88% |
| GSO | — | 6.9% |
| AlgoTune | — | 1.67 |
Agentic & Tool Use Not comparable
Gemma 4 31B IT: —, GPT-5: 33.1 (#56)
| Benchmark | Gemma 4 31B IT | GPT-5 |
|---|---|---|
| Terminal-Bench | — | 49.6% |
| GDPval | — | 34.8% |
| Remote Labor Index | — | 1.7% |
| DeepResearch Bench | — | 49.6% |
| BALROG | — | 32.8% |
| LMArena Search | — | 1133 |
| METR Time Horizons | — | 69.6% |
Reasoning GPT-5 leads
Gemma 4 31B IT: 27.2 (#122), GPT-5: 38.3 (#64)
| Benchmark | Gemma 4 31B IT | GPT-5 |
|---|---|---|
| Kagi LLM Benchmark | 63.5% | 72.7% |
| CritPt | 1.4% | 12.6% |
| Chess Puzzles | 5% | 37% |
| LMArena Hard Prompts | 1448 | 1416 |
| DTBench | 82.7% | 90.7% |
| LMCA | 39.3% | 40% |
| Epoch Capabilities Index | 142.74 | 150 |
| ARC-AGI-2 | — | 9.9% |
| SimpleBench | — | 56.7% |
| NYT Connections (extended) | 70.6% | — |
| ARC-AGI-1 | — | 65.7% |
| EnigmaEval | — | 10.5% |
| Thematic Generalization | 53% | — |
| EBR-Bench | — | 12.7% |
| Mystery Game Puzzles | — | 23% |
| Surface Evolver Bench | 30.6% | — |
| ForecastBench | — | 61.4 |
Math GPT-5 leads
Gemma 4 31B IT: 43.2 (#81), GPT-5: 55.0 (#44)
| Benchmark | Gemma 4 31B IT | GPT-5 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 73.3% | 91.4% |
| LMArena Math | 1465 | 1407 |
| FrontierMath (Tiers 1-3) | — | 55.4% |
| FrontierMath Tier 4 | — | 22% |
| ProofBench | — | 18% |
| Omni-MATH | — | 64.7% |
| MATH Level 5 | — | 98.1% |
| FrontierMath (Feb 2025 set) | — | 32.4% |
| FrontierMath Tier 4 (v1) | — | 12.5% |
Knowledge GPT-5 leads
Gemma 4 31B IT: 37.9 (#151), GPT-5: 56.6 (#43)
| Benchmark | Gemma 4 31B IT | GPT-5 |
|---|---|---|
| GPQA Diamond | 75.8% | 86.2% |
| SimpleQA Verified | 10.4% | 50.1% |
| Vectara Hallucination Rate | 7.4% | 14.7% |
| LMArena Expert | 1465 | 1419 |
| Humanity's Last Exam | — | 25.3% |
| MMLU-Pro | — | 86.3% |
| Confabulations | — | 10.3% |
| GPQA (HELM) | — | 79.2% |
Multimodal GPT-5 leads
Gemma 4 31B IT: 41.6 (#34), GPT-5: 46.8 (#13)
| Benchmark | Gemma 4 31B IT | GPT-5 |
|---|---|---|
| LMArena Vision | 1277 | 1232 |
| GeoBench | — | 81% |
| VPCT | — | 66% |
| LMArena Document | 1425 | — |
Multilingual Gemma 4 31B IT leads
Gemma 4 31B IT: 53.8 (#57), GPT-5: 51.4 (#110)
| Benchmark | Gemma 4 31B IT | GPT-5 |
|---|---|---|
| LMArena Non-English | 1431 | 1397 |
| LMArena Chinese | 1476 | 1422 |
| LMArena French | 1435 | 1410 |
| LMArena Russian | 1460 | 1406 |
| LMArena Spanish | 1444 | 1399 |
| LMArena German | — | 1416 |
| LMArena Japanese | — | 1409 |
| LMArena Korean | — | 1360 |
Instruction Following Gemma 4 31B IT leads
Gemma 4 31B IT: 75.5 (#61), GPT-5: 73.8 (#113)
| Benchmark | Gemma 4 31B IT | GPT-5 |
|---|---|---|
| LMArena Instruction Following | 1433 | 1388 |
| IFEval | — | 87.5% |
Long Context GPT-5 leads
Gemma 4 31B IT: 44.2 (#71), GPT-5: 69.5 (#2)
| Benchmark | Gemma 4 31B IT | GPT-5 |
|---|---|---|
| LMArena Longer Query | 1446 | 1399 |
| Fiction.LiveBench | — | 97.2% |
Writing & Preference GPT-5 leads
Gemma 4 31B IT: 60.5 (#96), GPT-5: 63.4 (#65)
| Benchmark | Gemma 4 31B IT | GPT-5 |
|---|---|---|
| LMArena Text | 1443 | 1406 |
| LMArena Creative Writing | 1415 | 1365 |
| EQ-Bench Creative Writing | 1368 | 1627 |
| LMArena Multi-Turn | 1452 | 1426 |
| Short-Story Creative Writing | — | 86% |
| WildBench | — | 85.7% |
| EQ-Bench 4 | 1120 | — |
Frequently asked questions
Is Gemma 4 31B IT better than GPT-5?
GPT-5 is the stronger model overall, scoring 50.9 to 43.5 on the Noometry Index. Gemma 4 31B IT costs 23× less per token, which makes it the better buy when GPT-5's lead doesn't matter for your workload.
Which is cheaper, Gemma 4 31B IT or GPT-5?
Gemma 4 31B IT is cheaper. It lists at $0.09 per million input tokens and $0.34 per million output tokens; GPT-5 lists at $1.25 and $10.
Is Gemma 4 31B IT or GPT-5 better for coding?
GPT-5 scores higher on coding benchmarks: 50.3 versus 42.3 in the Noometry coding category.
Which has the bigger context window?
GPT-5 does, with 400K tokens against 262K.
How many benchmarks do Gemma 4 31B IT and GPT-5 share?
30 benchmarks have published results for both models. Gemma 4 31B IT has 35 scored results on Noometry and GPT-5 has 69.