Model comparison
GPT-4.5 vs GPT-5.6 Terra
GPT-5.6 Terra is the stronger model overall, scoring 59.1 to 36.9 on the Noometry Index.
Last verified . 25 shared benchmarks.
Summary
- They share 25 benchmarks with published results for both. GPT-4.5 scores higher in 0 categories and GPT-5.6 Terra in 10 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Terra leads 81.5 to 31.0.
- The biggest single-benchmark swing is ARC-AGI-1: 10.3% for GPT-4.5 and 96.5% for GPT-5.6 Terra.
Side by side
| GPT-4.5 | GPT-5.6 Terra | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 36.9 | 59.1 |
| Released | 2025-02-27 | 2026-07-09 |
| Weights | Proprietary | Proprietary |
| Context window | — | 1.05M |
| Max output | — | 128K |
| Input $ / M tokens | — | $2 |
| Output $ / M tokens | — | $12 |
| Results tracked | 42 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.6 Terra leads
GPT-4.5: 41.8 (#102), GPT-5.6 Terra: 57.1 (#16)
| Benchmark | GPT-4.5 | GPT-5.6 Terra |
|---|---|---|
| WeirdML | 39.4% | 78.3% |
| LMArena Coding | 1397 | 1484 |
| DeepSWE | — | 69.6% |
| FrontierCode | — | 41.3% |
| Aider Polyglot | 44.9% | — |
| CursorBench | — | 41.3% |
| LMArena WebDev | — | 1520 |
| SciCode | — | 55% |
| LiveBench Coding | 75.2% | — |
| ALE-Bench | — | 1,951 |
Agentic & Tool Use GPT-5.6 Terra leads
GPT-4.5: 27.6 (#92), GPT-5.6 Terra: 40.2 (#23)
| Benchmark | GPT-4.5 | GPT-5.6 Terra |
|---|---|---|
| APEX-Agents | — | 58.2% |
| Cybench | 17.5% | — |
| BALROG | — | 53.2% |
| GDP.pdf | — | 24.7% |
| Vending-Bench 2 | — | 7,343 |
Reasoning GPT-5.6 Terra leads
GPT-4.5: 14.1 (#320), GPT-5.6 Terra: 60.6 (#19)
| Benchmark | GPT-4.5 | GPT-5.6 Terra |
|---|---|---|
| ARC-AGI-2 | 0.8% | 83.9% |
| SimpleBench | 34.5% | 48.9% |
| ARC-AGI-1 | 10.3% | 96.5% |
| LMArena Hard Prompts | 1404 | 1467 |
| Epoch Capabilities Index | 136.74 | 159.31 |
| Kagi LLM Benchmark | — | 51.3% |
| NYT Connections (extended) | — | 78.4% |
| CritPt | — | 30% |
| Chess Puzzles | — | 54% |
| EnigmaEval | 3.2% | — |
| LiveBench Reasoning | 71.1% | — |
| Mystery Game Puzzles | — | 35% |
| DTBench | — | 93.3% |
| LiveBench Data Analysis | 64.3% | — |
| LMCA | — | 55% |
| Surface Evolver Bench | — | 83.8% |
| ForecastBench | 61.7 | — |
| LiveBench | 69% | — |
Math GPT-5.6 Terra leads
GPT-4.5: 31.0 (#200), GPT-5.6 Terra: 81.5 (#10)
| Benchmark | GPT-4.5 | GPT-5.6 Terra |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 37.8% | 99.7% |
| LMArena Math | 1412 | 1463 |
| FrontierMath (Tiers 1-3) | — | 86% |
| FrontierMath Tier 4 | — | 70.7% |
| ProofBench | — | 74% |
| LiveBench Math | 69.3% | — |
| MATH Level 5 | 78.6% | — |
Knowledge GPT-5.6 Terra leads
GPT-4.5: 32.4 (#205), GPT-5.6 Terra: 61.3 (#27)
| Benchmark | GPT-4.5 | GPT-5.6 Terra |
|---|---|---|
| GPQA Diamond | 68.7% | 93.3% |
| LMArena Expert | 1394 | 1492 |
| Humanity's Last Exam | 5.4% | — |
| SimpleQA Verified | — | 43.2% |
| Confabulations | 13.6% | — |
Multimodal GPT-5.6 Terra leads
GPT-4.5: 37.0 (#68), GPT-5.6 Terra: 47.5 (#10)
| Benchmark | GPT-4.5 | GPT-5.6 Terra |
|---|---|---|
| LMArena Vision | 1195 | 1271 |
| VPCT | 45% | — |
| Blueprint-Bench 2 | — | 30.8% |
| Furniture Assembly | — | 54.2% |
| LMArena Document | — | 1472 |
Multilingual GPT-5.6 Terra leads
GPT-4.5: 52.7 (#78), GPT-5.6 Terra: 54.5 (#42)
| Benchmark | GPT-4.5 | GPT-5.6 Terra |
|---|---|---|
| LMArena Non-English | 1413 | 1439 |
| LMArena Chinese | 1421 | 1514 |
| LMArena French | 1417 | 1476 |
| LMArena German | 1457 | 1462 |
| LMArena Japanese | 1416 | 1446 |
| LMArena Korean | 1392 | 1420 |
| LMArena Russian | 1419 | 1450 |
| LMArena Spanish | — | 1455 |
Instruction Following GPT-5.6 Terra leads
GPT-4.5: 72.8 (#129), GPT-5.6 Terra: 76.6 (#38)
| Benchmark | GPT-4.5 | GPT-5.6 Terra |
|---|---|---|
| LMArena Instruction Following | 1404 | 1453 |
| LiveBench Instruction Following | 72.3% | — |
Long Context GPT-5.6 Terra leads
GPT-4.5: 40.4 (#150), GPT-5.6 Terra: 44.4 (#64)
| Benchmark | GPT-4.5 | GPT-5.6 Terra |
|---|---|---|
| LMArena Longer Query | 1406 | 1452 |
| Fiction.LiveBench | 63.9% | — |
Writing & Preference GPT-5.6 Terra leads
GPT-4.5: 57.1 (#129), GPT-5.6 Terra: 70.4 (#21)
| Benchmark | GPT-4.5 | GPT-5.6 Terra |
|---|---|---|
| LMArena Text | 1417 | 1446 |
| LMArena Creative Writing | 1394 | 1411 |
| EQ-Bench Creative Writing | 1258 | 1855 |
| LMArena Multi-Turn | 1444 | 1451 |
| Short-Story Creative Writing | 75.6% | — |
| EQ-Bench 4 | — | 1234 |
| LiveBench Language | 61.5% | — |
Frequently asked questions
Is GPT-4.5 better than GPT-5.6 Terra?
GPT-5.6 Terra is the stronger model overall, scoring 59.1 to 36.9 on the Noometry Index.
Is GPT-4.5 or GPT-5.6 Terra better for coding?
GPT-5.6 Terra scores higher on coding benchmarks: 57.1 versus 41.8 in the Noometry coding category.
How many benchmarks do GPT-4.5 and GPT-5.6 Terra share?
25 benchmarks have published results for both models. GPT-4.5 has 42 scored results on Noometry and GPT-5.6 Terra has 52.