# Gemma 3n E4b IT vs GPT-5.6 Terra

> GPT-5.6 Terra is the stronger model overall, scoring 59.1 to 36.7 on the Noometry Index. Gemma 3n E4b IT costs 60× less per token, which makes it the better buy when GPT-5.6 Terra's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gemma-3n-e4b-it-vs-gpt-5-6-terra
- Last updated: 2026-10-10
- Shared benchmarks: 18

## Summary

- They share 18 benchmarks with published results for both. Gemma 3n E4b IT scores higher in 0 categories and GPT-5.6 Terra in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Terra leads 81.5 to 32.5.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 31.5% for Gemma 3n E4b IT and 51.3% for GPT-5.6 Terra.
- Gemma 3n E4b IT is cheaper at $0.06 / $0.12 per million input/output tokens, against $2 / $12 for GPT-5.6 Terra.
- GPT-5.6 Terra accepts more context: 1.05M tokens versus 33K.
- Gemma 3n E4b IT has downloadable open weights; the other is API-only.

## Snapshot

| | Gemma 3n E4b IT | GPT-5.6 Terra |
|---|---|---|
| Provider | Google | OpenAI |
| Noometry Index | 36.7 | 59.1 |
| Rank | 203 | 15 |
| Context | 33K | 1.05M |
| Input $/M | $0.06 | $2 |
| Output $/M | $0.12 | $12 |
| Weights | Open | Proprietary |

## Coding

- Gemma 3n E4b IT: 36.4 (#193)
- GPT-5.6 Terra: 57.1 (#16)

| Benchmark | Gemma 3n E4b IT | GPT-5.6 Terra |
|---|---|---|
| LMArena Coding | 1269 | 1484 |
| DeepSWE | — | 69.6% |
| FrontierCode | — | 41.3% |
| CursorBench | — | 41.3% |
| LMArena WebDev | — | 1520 |
| SciCode | — | 55% |
| WeirdML | — | 78.3% |
| ALE-Bench | — | 1,951 |

## Agentic & Tool Use

- Gemma 3n E4b IT: —
- GPT-5.6 Terra: 40.2 (#23)

| Benchmark | Gemma 3n E4b IT | GPT-5.6 Terra |
|---|---|---|
| APEX-Agents | — | 58.2% |
| BALROG | — | 53.2% |
| GDP.pdf | — | 24.7% |
| Vending-Bench 2 | — | 7,343 |

## Reasoning

- Gemma 3n E4b IT: 19.2 (#239)
- GPT-5.6 Terra: 60.6 (#19)

| Benchmark | Gemma 3n E4b IT | GPT-5.6 Terra |
|---|---|---|
| Kagi LLM Benchmark | 31.5% | 51.3% |
| LMArena Hard Prompts | 1284 | 1467 |
| ARC-AGI-2 | — | 83.9% |
| SimpleBench | — | 48.9% |
| NYT Connections (extended) | — | 78.4% |
| ARC-AGI-1 | — | 96.5% |
| CritPt | — | 30% |
| Chess Puzzles | — | 54% |
| Mystery Game Puzzles | — | 35% |
| DTBench | — | 93.3% |
| LMCA | — | 55% |
| Surface Evolver Bench | — | 83.8% |
| Epoch Capabilities Index | — | 159.31 |

## Math

- Gemma 3n E4b IT: 32.5 (#186)
- GPT-5.6 Terra: 81.5 (#10)

| Benchmark | Gemma 3n E4b IT | GPT-5.6 Terra |
|---|---|---|
| LMArena Math | 1251 | 1463 |
| FrontierMath (Tiers 1-3) | — | 86% |
| FrontierMath Tier 4 | — | 70.7% |
| OTIS Mock AIME 2024-2025 | — | 99.7% |
| ProofBench | — | 74% |

## Knowledge

- Gemma 3n E4b IT: 33.6 (#193)
- GPT-5.6 Terra: 61.3 (#27)

| Benchmark | Gemma 3n E4b IT | GPT-5.6 Terra |
|---|---|---|
| LMArena Expert | 1245 | 1492 |
| GPQA Diamond | — | 93.3% |
| SimpleQA Verified | — | 43.2% |

## Multimodal

- Gemma 3n E4b IT: —
- GPT-5.6 Terra: 47.5 (#10)

| Benchmark | Gemma 3n E4b IT | GPT-5.6 Terra |
|---|---|---|
| LMArena Vision | — | 1271 |
| Blueprint-Bench 2 | — | 30.8% |
| Furniture Assembly | — | 54.2% |
| LMArena Document | — | 1472 |

## Multilingual

- Gemma 3n E4b IT: 43.5 (#179)
- GPT-5.6 Terra: 54.5 (#42)

| Benchmark | Gemma 3n E4b IT | GPT-5.6 Terra |
|---|---|---|
| LMArena Non-English | 1285 | 1439 |
| LMArena Chinese | 1310 | 1514 |
| LMArena French | 1330 | 1476 |
| LMArena German | 1312 | 1462 |
| LMArena Japanese | 1274 | 1446 |
| LMArena Korean | 1259 | 1420 |
| LMArena Russian | 1288 | 1450 |
| LMArena Spanish | 1305 | 1455 |

## Instruction Following

- Gemma 3n E4b IT: 66.3 (#205)
- GPT-5.6 Terra: 76.6 (#38)

| Benchmark | Gemma 3n E4b IT | GPT-5.6 Terra |
|---|---|---|
| LMArena Instruction Following | 1255 | 1453 |

## Long Context

- Gemma 3n E4b IT: 38.8 (#185)
- GPT-5.6 Terra: 44.4 (#64)

| Benchmark | Gemma 3n E4b IT | GPT-5.6 Terra |
|---|---|---|
| LMArena Longer Query | 1277 | 1452 |

## Writing & Preference

- Gemma 3n E4b IT: 50.3 (#181)
- GPT-5.6 Terra: 70.4 (#21)

| Benchmark | Gemma 3n E4b IT | GPT-5.6 Terra |
|---|---|---|
| LMArena Text | 1306 | 1446 |
| LMArena Creative Writing | 1287 | 1411 |
| LMArena Multi-Turn | 1276 | 1451 |
| EQ-Bench Creative Writing | — | 1855 |
| EQ-Bench 4 | — | 1234 |

## FAQ

### Is Gemma 3n E4b IT better than GPT-5.6 Terra?

GPT-5.6 Terra is the stronger model overall, scoring 59.1 to 36.7 on the Noometry Index. Gemma 3n E4b IT costs 60× less per token, which makes it the better buy when GPT-5.6 Terra's lead doesn't matter for your workload.

### Which is cheaper, Gemma 3n E4b IT or GPT-5.6 Terra?

Gemma 3n E4b IT is cheaper. It lists at $0.06 per million input tokens and $0.12 per million output tokens; GPT-5.6 Terra lists at $2 and $12.

### Is Gemma 3n E4b IT or GPT-5.6 Terra better for coding?

GPT-5.6 Terra scores higher on coding benchmarks: 57.1 versus 36.4 in the Noometry coding category.

### Which has the bigger context window?

GPT-5.6 Terra does, with 1.05M tokens against 33K.

### How many benchmarks do Gemma 3n E4b IT and GPT-5.6 Terra share?

18 benchmarks have published results for both models. Gemma 3n E4b IT has 18 scored results on Noometry and GPT-5.6 Terra has 52.