Model comparison
GPT-4 vs GPT-5.6 Terra
GPT-5.6 Terra is the stronger model overall, scoring 59.2 to 29.1 on the Noometry Index.
Last verified . 26 shared benchmarks.
Summary
- They share 26 benchmarks with published results for both. GPT-4 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 81.6 to 10.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 1.1% for GPT-4 and 99.7% for GPT-5.6 Terra.
- GPT-5.6 Terra is cheaper at $2 / $12 per million input/output tokens, against $30 / $60 for GPT-4.
- GPT-5.6 Terra accepts more context: 1.05M tokens versus 8K.
Side by side
| GPT-4 | GPT-5.6 Terra | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 29.1 | 59.2 |
| Released | 2023-03-14 | 2026-07-09 |
| Weights | Proprietary | Proprietary |
| Context window | 8K | 1.05M |
| Max output | 8K | 128K |
| Input $ / M tokens | $30 | $2 |
| Output $ / M tokens | $60 | $12 |
| Results tracked | 38 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.6 Terra leads
GPT-4: 31.6 (#283), GPT-5.6 Terra: 57.7 (#19)
| Benchmark | GPT-4 | GPT-5.6 Terra |
|---|---|---|
| WeirdML | 12.4% | 78.3% |
| LMArena Coding | 1254 | 1484 |
| DeepSWE | — | 69.6% |
| FrontierCode | — | 41.3% |
| CursorBench | — | 41.3% |
| LMArena WebDev | — | 1522 |
| SciCode | — | 55% |
| BigCodeBench Instruct | 46% | — |
| BigCodeBench Complete | 57.2% | — |
| ALE-Bench | — | 1,951 |
| HumanEval+ | 79.3% | — |
Agentic & Tool Use Not comparable
GPT-4: —, GPT-5.6 Terra: 40.1 (#25)
| Benchmark | GPT-4 | GPT-5.6 Terra |
|---|---|---|
| APEX-Agents | — | 58.2% |
| BALROG | — | 53.2% |
| GDP.pdf | — | 24.7% |
| METR Time Horizons | 36.1% | — |
| Vending-Bench 2 | — | 7,343 |
Reasoning GPT-5.6 Terra leads
GPT-4: 17.8 (#289), GPT-5.6 Terra: 60.7 (#21)
| Benchmark | GPT-4 | GPT-5.6 Terra |
|---|---|---|
| Chess Puzzles | 4% | 54% |
| LMArena Hard Prompts | 1241 | 1468 |
| Mystery Game Puzzles | 12% | 35% |
| DTBench | 62.7% | 93.3% |
| LMCA | 17.1% | 55% |
| Epoch Capabilities Index | 125.89 | 159.62 |
| ARC-AGI-2 | — | 83.9% |
| SimpleBench | — | 48.9% |
| Kagi LLM Benchmark | — | 51.3% |
| NYT Connections (extended) | — | 78.4% |
| ARC-AGI-1 | — | 96.5% |
| CritPt | — | 30% |
| Surface Evolver Bench | — | 83.8% |
| BIG-Bench Hard | 75.1% | — |
| ForecastBench | 57.8 | — |
| HellaSwag | 95.3% | — |
| WinoGrande | 87.5% | — |
Math GPT-5.6 Terra leads
GPT-4: 10.8 (#309), GPT-5.6 Terra: 81.6 (#12)
| Benchmark | GPT-4 | GPT-5.6 Terra |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.1% | 99.7% |
| LMArena Math | 1269 | 1466 |
| FrontierMath (Tiers 1-3) | — | 86% |
| FrontierMath Tier 4 | — | 70.7% |
| ProofBench | — | 74% |
| MATH Level 5 | 23% | — |
| GSM8K | 92% | — |
Knowledge GPT-5.6 Terra leads
GPT-4: 18.4 (#282), GPT-5.6 Terra: 61.2 (#30)
| Benchmark | GPT-4 | GPT-5.6 Terra |
|---|---|---|
| GPQA Diamond | 35.7% | 93.3% |
| LMArena Expert | 1211 | 1492 |
| SimpleQA Verified | — | 43.2% |
| MMLU | 86.4% | — |
| TriviaQA | 84.8% | — |
Multimodal Not comparable
GPT-4: —, GPT-5.6 Terra: 47.3 (#11)
| Benchmark | GPT-4 | GPT-5.6 Terra |
|---|---|---|
| LMArena Vision | — | 1271 |
| Blueprint-Bench 2 | — | 30.8% |
| Furniture Assembly | — | 54.2% |
| LMArena Document | — | 1472 |
Multilingual GPT-5.6 Terra leads
GPT-4: 40.6 (#215), GPT-5.6 Terra: 54.4 (#44)
| Benchmark | GPT-4 | GPT-5.6 Terra |
|---|---|---|
| LMArena Non-English | 1246 | 1439 |
| LMArena Chinese | 1242 | 1513 |
| LMArena French | 1283 | 1471 |
| LMArena German | 1251 | 1460 |
| LMArena Japanese | 1209 | 1457 |
| LMArena Korean | 1184 | 1425 |
| LMArena Russian | 1251 | 1450 |
| LMArena Spanish | 1261 | 1448 |
Instruction Following GPT-5.6 Terra leads
GPT-4: 65.3 (#222), GPT-5.6 Terra: 76.4 (#40)
| Benchmark | GPT-4 | GPT-5.6 Terra |
|---|---|---|
| LMArena Instruction Following | 1241 | 1454 |
Long Context GPT-5.6 Terra leads
GPT-4: 37.7 (#212), GPT-5.6 Terra: 44.4 (#68)
| Benchmark | GPT-4 | GPT-5.6 Terra |
|---|---|---|
| LMArena Longer Query | 1244 | 1451 |
Writing & Preference GPT-5.6 Terra leads
GPT-4: 34.9 (#268), GPT-5.6 Terra: 70.2 (#23)
| Benchmark | GPT-4 | GPT-5.6 Terra |
|---|---|---|
| LMArena Text | 1263 | 1447 |
| LMArena Creative Writing | 1244 | 1410 |
| EQ-Bench Creative Writing | 752 | 1855 |
| LMArena Multi-Turn | 1257 | 1449 |
| EQ-Bench 4 | — | 1234 |
Frequently asked questions
Is GPT-4 better than GPT-5.6 Terra?
GPT-5.6 Terra is the stronger model overall, scoring 59.2 to 29.1 on the Noometry Index.
Which is cheaper, GPT-4 or GPT-5.6 Terra?
GPT-5.6 Terra is cheaper. It lists at $2 per million input tokens and $12 per million output tokens; GPT-4 lists at $30 and $60.
Is GPT-4 or GPT-5.6 Terra better for coding?
GPT-5.6 Terra scores higher on coding benchmarks: 57.7 versus 31.6 in the Noometry coding category.
Which has the bigger context window?
GPT-5.6 Terra does, with 1.05M tokens against 8K.
How many benchmarks do GPT-4 and GPT-5.6 Terra share?
26 benchmarks have published results for both models. GPT-4 has 38 scored results on Noometry and GPT-5.6 Terra has 52.