Model comparison
Gemma 4 31B IT vs GPT-4.5
Gemma 4 31B IT is the stronger model overall, scoring 43.5 to 37.2 on the Noometry Index.
Last verified . 19 shared benchmarks.
GoogleSummary
- They share 19 benchmarks with published results for both. Gemma 4 31B IT scores higher in 9 categories and GPT-4.5 in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemma 4 31B IT leads 27.2 to 13.9.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 73.3% for Gemma 4 31B IT and 37.8% for GPT-4.5.
- Gemma 4 31B IT has downloadable open weights; the other is API-only.
Side by side
| Gemma 4 31B IT | GPT-4.5 | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 43.5 | 37.2 |
| Released | 2026-04-02 | 2025-02-27 |
| Weights | Open | Proprietary |
| Context window | 262K | — |
| Max output | 33K | — |
| Input $ / M tokens | $0.09 | — |
| Output $ / M tokens | $0.34 | — |
| Results tracked | 35 | 42 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Gemma 4 31B IT: 42.3 (#108), GPT-4.5: 42.2 (#109)
| Benchmark | Gemma 4 31B IT | GPT-4.5 |
|---|---|---|
| WeirdML | 52.3% | 39.4% |
| LMArena Coding | 1459 | 1396 |
| Aider Polyglot | — | 44.9% |
| LMArena WebDev | 1366 | — |
| SciCode | 43.4% | — |
| LiveBench Coding | — | 75.2% |
| ALE-Bench | 925.5 | — |
Agentic & Tool Use Not comparable
Gemma 4 31B IT: —, GPT-4.5: 27.9 (#97)
| Benchmark | Gemma 4 31B IT | GPT-4.5 |
|---|---|---|
| Cybench | — | 17.5% |
Reasoning Gemma 4 31B IT leads
Gemma 4 31B IT: 27.2 (#122), GPT-4.5: 13.9 (#330)
| Benchmark | Gemma 4 31B IT | GPT-4.5 |
|---|---|---|
| LMArena Hard Prompts | 1448 | 1403 |
| Epoch Capabilities Index | 142.74 | 136.74 |
| ARC-AGI-2 | — | 0.8% |
| SimpleBench | — | 34.5% |
| Kagi LLM Benchmark | 63.5% | — |
| NYT Connections (extended) | 70.6% | — |
| ARC-AGI-1 | — | 10.3% |
| CritPt | 1.4% | — |
| Chess Puzzles | 5% | — |
| EnigmaEval | — | 3.2% |
| Thematic Generalization | 53% | — |
| LiveBench Reasoning | — | 71.1% |
| DTBench | 82.7% | — |
| LiveBench Data Analysis | — | 64.3% |
| LMCA | 39.3% | — |
| Surface Evolver Bench | 30.6% | — |
| ForecastBench | — | 61.7 |
| LiveBench | — | 69% |
Math Gemma 4 31B IT leads
Gemma 4 31B IT: 43.2 (#81), GPT-4.5: 32.6 (#211)
| Benchmark | Gemma 4 31B IT | GPT-4.5 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 73.3% | 37.8% |
| LMArena Math | 1465 | 1412 |
| LiveBench Math | — | 69.3% |
| MATH Level 5 | — | 78.6% |
Knowledge Gemma 4 31B IT leads
Gemma 4 31B IT: 37.9 (#151), GPT-4.5: 32.5 (#211)
| Benchmark | Gemma 4 31B IT | GPT-4.5 |
|---|---|---|
| GPQA Diamond | 75.8% | 68.7% |
| LMArena Expert | 1465 | 1394 |
| Humanity's Last Exam | — | 5.4% |
| SimpleQA Verified | 10.4% | — |
| Confabulations | — | 13.6% |
| Vectara Hallucination Rate | 7.4% | — |
Multimodal Gemma 4 31B IT leads
Gemma 4 31B IT: 41.6 (#34), GPT-4.5: 37.6 (#71)
| Benchmark | Gemma 4 31B IT | GPT-4.5 |
|---|---|---|
| LMArena Vision | 1277 | 1195 |
| VPCT | — | 45% |
| LMArena Document | 1425 | — |
Multilingual Gemma 4 31B IT leads
Gemma 4 31B IT: 53.8 (#57), GPT-4.5: 52.5 (#83)
| Benchmark | Gemma 4 31B IT | GPT-4.5 |
|---|---|---|
| LMArena Non-English | 1431 | 1413 |
| LMArena Chinese | 1476 | 1421 |
| LMArena French | 1435 | 1418 |
| LMArena Russian | 1460 | 1419 |
| LMArena German | — | 1457 |
| LMArena Japanese | — | 1416 |
| LMArena Korean | — | 1392 |
| LMArena Spanish | 1444 | — |
Instruction Following Gemma 4 31B IT leads
Gemma 4 31B IT: 75.5 (#61), GPT-4.5: 72.6 (#134)
| Benchmark | Gemma 4 31B IT | GPT-4.5 |
|---|---|---|
| LMArena Instruction Following | 1433 | 1404 |
| LiveBench Instruction Following | — | 72.3% |
Long Context Gemma 4 31B IT leads
Gemma 4 31B IT: 44.2 (#71), GPT-4.5: 40.4 (#155)
| Benchmark | Gemma 4 31B IT | GPT-4.5 |
|---|---|---|
| LMArena Longer Query | 1446 | 1406 |
| Fiction.LiveBench | — | 63.9% |
Writing & Preference Gemma 4 31B IT leads
Gemma 4 31B IT: 60.5 (#96), GPT-4.5: 56.9 (#134)
| Benchmark | Gemma 4 31B IT | GPT-4.5 |
|---|---|---|
| LMArena Text | 1443 | 1417 |
| LMArena Creative Writing | 1415 | 1394 |
| EQ-Bench Creative Writing | 1368 | 1258 |
| LMArena Multi-Turn | 1452 | 1444 |
| Short-Story Creative Writing | — | 75.6% |
| EQ-Bench 4 | 1120 | — |
| LiveBench Language | — | 61.5% |
Frequently asked questions
Is Gemma 4 31B IT better than GPT-4.5?
Gemma 4 31B IT is the stronger model overall, scoring 43.5 to 37.2 on the Noometry Index.
Is Gemma 4 31B IT or GPT-4.5 better for coding?
They score almost the same on coding (42.3 vs 42.2); test both on your own repository before choosing.
How many benchmarks do Gemma 4 31B IT and GPT-4.5 share?
19 benchmarks have published results for both models. Gemma 4 31B IT has 35 scored results on Noometry and GPT-4.5 has 42.