Gemma 4 31B IT
| Released | 2026-04-02 |
|---|---|
| Price per 1M | $0.14 in · $0.40 out |
| Overall score | 47.9 · rank 131 |
| Context window | 262,144 tokens |
| Provider | |
| Weights | Open weights |
Scores
| Dimension | Score | Rank | Benchmarks |
|---|---|---|---|
| Overall: Balanced | 47.9 | 131 | 21 |
| Overall: Coding | 46.6 | 139 | 13 |
| Overall: Reasoning | 47.7 | 134 | 8 |
| Overall: Agentic work | 48.4 | 124 | 11 |
| Overall: Frontend | 45.8 | 129 | 5 |
| Coding | 45.7 | 149 | 2 |
| Agents | 49.1 | 111 | 3 |
| Reasoning | 50.5 | 111 | 6 |
| Math | 49.2 | 120 | 1 |
| Knowledge | 38.7 | 146 | 1 |
| Frontend | 43.9 | 97 | 2 |
| General | 52.1 | 87 | 5 |
| Multimodal | 57.6 | 31 | 1 |
| Value vs price | +4.8 | 78 |
Configurations
| Configuration | Balanced | Benchmarks |
|---|---|---|
| Unspecified setting | 47.0 ±1.3 | 18 |
| Minimal | 48.8 ±5.6 | 2 |
| Thinking | 55.1 ±5.6 | 2 |
Evidence
Coding 45.7 ±2.0
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LiveBench 2.0 Coding | 20.3% | LiveBench | 2026-11-06 |
| SciCode | 45.5% | Epoch AI | 2026-11-11 |
Agents 49.1 ±3.2
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LiveBench 2.0 Agentic | 8.5% | LiveBench | 2026-11-06 |
| LiveBench 2.0 Browser Use | 0.0% | LiveBench | 2026-11-06 |
| LiveBench 2.0 Research | 38.1% | LiveBench | 2026-11-06 |
Reasoning 50.5 ±4.5
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Extended NYT Connections | 70.6% | Lech Mazur | 2026-11-10 |
| GPQA Diamond | 75.8% | Epoch AI | 2026-08-06 |
| CritPt | 1.4% | Epoch AI | 2026-11-11 |
| Kagi Reasoning Benchmark | 63.5% | Kagi | 2026-11-11 |
| LiveBench 2.0 Reasoning | 68.3% | LiveBench | 2026-11-06 |
| LiveBench 2.0 Data Analysis | 53.3% | LiveBench | 2026-11-06 |
Math 49.2 ±3.6
| Benchmark | Result | Source | Date |
|---|---|---|---|
| OTIS Mock AIME 2024-2025 | 73.3% | Epoch AI | 2026-08-06 |
Knowledge 38.7 ±3.0
| Benchmark | Result | Source | Date |
|---|---|---|---|
| SimpleQA Verified | 10.4% | Epoch AI | 2026-08-27 |
Frontend 43.9 ±2.0
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LiveBench 2.0 Design | 56.5% | LiveBench | 2026-11-06 |
| LMArena WebDev | 1,370 | LMArena | 2026-11-10 |
General 52.1 ±2.6
| Benchmark | Result | Source | Date |
|---|---|---|---|
| EQ-Bench 4 | 1,110 | EQ-Bench | 2026-10-24 |
| EQ-Bench Creative Writing v3 | 1,369 | EQ-Bench | 2026-11-11 |
| LMArena Text | 1,452 | LMArena | 2026-11-09 |
| Creative Story-Writing (Lech Mazur) | 23.0% | Lech Mazur | 2026-10-23 |
| EQ-Bench Longform Creative Writing | 56.5% | EQ-Bench | 2026-11-11 |
Multimodal 57.6 ±4.1
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LMArena Vision | 1,265 | LMArena | 2026-11-09 |
Results not used in the score
| Benchmark | Result | Source | Why not used |
|---|---|---|---|
| Extended NYT Connections | 15.7% | Lech Mazur | A stronger configuration result was used |
Names matched to this model
gemma-4-31b, gemma-4-31b-it