Model comparison
GPT-5.6 Terra vs GPT-5 Pro
GPT-5.6 Terra is the stronger model overall, scoring 59.2 to 46.4 on the Noometry Index.
Last verified . 8 shared benchmarks.
Summary
- They share 8 benchmarks with published results for both. GPT-5.6 Terra scores higher in 4 categories and GPT-5 Pro in 0 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Terra leads 81.6 to 48.5.
- The biggest single-benchmark swing is ARC-AGI-2: 83.9% for GPT-5.6 Terra and 18.3% for GPT-5 Pro.
- GPT-5.6 Terra is cheaper at $2 / $12 per million input/output tokens, against $15 / $120 for GPT-5 Pro.
- GPT-5.6 Terra accepts more context: 1.05M tokens versus 400K.
Side by side
| GPT-5.6 Terra | GPT-5 Pro | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 59.2 | 46.4 |
| Released | 2026-07-09 | 2025-10-06 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 400K |
| Max output | 128K | 272K |
| Input $ / M tokens | $2 | $15 |
| Output $ / M tokens | $12 | $120 |
| Results tracked | 52 | 12 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.6 Terra leads
GPT-5.6 Terra: 57.7 (#19), GPT-5 Pro: 44.0 (#80)
| Benchmark | GPT-5.6 Terra | GPT-5 Pro |
|---|---|---|
| WeirdML | 78.3% | 60.4% |
| DeepSWE | 69.6% | — |
| FrontierCode | 41.3% | — |
| CursorBench | 41.3% | — |
| LMArena WebDev | 1522 | — |
| SciCode | 55% | — |
| LMArena Coding | 1484 | — |
| ALE-Bench | 1,951 | — |
| AlgoTune | — | 1.31 |
Agentic & Tool Use Not comparable
GPT-5.6 Terra: 40.1 (#25), GPT-5 Pro: —
| Benchmark | GPT-5.6 Terra | GPT-5 Pro |
|---|---|---|
| APEX-Agents | 58.2% | — |
| BALROG | 53.2% | — |
| GDP.pdf | 24.7% | — |
| Vending-Bench 2 | 7,343 | — |
Reasoning GPT-5.6 Terra leads
GPT-5.6 Terra: 60.7 (#21), GPT-5 Pro: 38.9 (#62)
| Benchmark | GPT-5.6 Terra | GPT-5 Pro |
|---|---|---|
| ARC-AGI-2 | 83.9% | 18.3% |
| SimpleBench | 48.9% | 61.6% |
| Kagi LLM Benchmark | 51.3% | 76.8% |
| ARC-AGI-1 | 96.5% | 70.2% |
| Epoch Capabilities Index | 159.62 | 150.28 |
| NYT Connections (extended) | 78.4% | — |
| CritPt | 30% | — |
| Chess Puzzles | 54% | — |
| EnigmaEval | — | 18.8% |
| LMArena Hard Prompts | 1468 | — |
| Mystery Game Puzzles | 35% | — |
| DTBench | 93.3% | — |
| LMCA | 55% | — |
| Surface Evolver Bench | 83.8% | — |
Math GPT-5.6 Terra leads
GPT-5.6 Terra: 81.6 (#12), GPT-5 Pro: 48.5 (#63)
| Benchmark | GPT-5.6 Terra | GPT-5 Pro |
|---|---|---|
| FrontierMath (Tiers 1-3) | 86% | 55.8% |
| FrontierMath Tier 4 | 70.7% | 19.5% |
| OTIS Mock AIME 2024-2025 | 99.7% | — |
| ProofBench | 74% | — |
| LMArena Math | 1466 | — |
| FrontierMath Tier 4 (v1) | — | 14.6% |
Knowledge GPT-5.6 Terra leads
GPT-5.6 Terra: 61.2 (#30), GPT-5 Pro: 56.7 (#42)
| Benchmark | GPT-5.6 Terra | GPT-5 Pro |
|---|---|---|
| GPQA Diamond | 93.3% | — |
| Humanity's Last Exam | — | 31.6% |
| SimpleQA Verified | 43.2% | — |
| LMArena Expert | 1492 | — |
Multimodal Not comparable
GPT-5.6 Terra: 47.3 (#11), GPT-5 Pro: —
| Benchmark | GPT-5.6 Terra | GPT-5 Pro |
|---|---|---|
| LMArena Vision | 1271 | — |
| Blueprint-Bench 2 | 30.8% | — |
| Furniture Assembly | 54.2% | — |
| LMArena Document | 1472 | — |
Multilingual Not comparable
GPT-5.6 Terra: 54.4 (#44), GPT-5 Pro: —
| Benchmark | GPT-5.6 Terra | GPT-5 Pro |
|---|---|---|
| LMArena Non-English | 1439 | — |
| LMArena Chinese | 1513 | — |
| LMArena French | 1471 | — |
| LMArena German | 1460 | — |
| LMArena Japanese | 1457 | — |
| LMArena Korean | 1425 | — |
| LMArena Russian | 1450 | — |
| LMArena Spanish | 1448 | — |
Instruction Following Not comparable
GPT-5.6 Terra: 76.4 (#40), GPT-5 Pro: —
| Benchmark | GPT-5.6 Terra | GPT-5 Pro |
|---|---|---|
| LMArena Instruction Following | 1454 | — |
Long Context Not comparable
GPT-5.6 Terra: 44.4 (#68), GPT-5 Pro: —
| Benchmark | GPT-5.6 Terra | GPT-5 Pro |
|---|---|---|
| LMArena Longer Query | 1451 | — |
Writing & Preference Not comparable
GPT-5.6 Terra: 70.2 (#23), GPT-5 Pro: —
| Benchmark | GPT-5.6 Terra | GPT-5 Pro |
|---|---|---|
| LMArena Text | 1447 | — |
| LMArena Creative Writing | 1410 | — |
| EQ-Bench Creative Writing | 1855 | — |
| EQ-Bench 4 | 1234 | — |
| LMArena Multi-Turn | 1449 | — |
Frequently asked questions
Is GPT-5.6 Terra better than GPT-5 Pro?
GPT-5.6 Terra is the stronger model overall, scoring 59.2 to 46.4 on the Noometry Index.
Which is cheaper, GPT-5.6 Terra or GPT-5 Pro?
GPT-5.6 Terra is cheaper. It lists at $2 per million input tokens and $12 per million output tokens; GPT-5 Pro lists at $15 and $120.
Is GPT-5.6 Terra or GPT-5 Pro better for coding?
GPT-5.6 Terra scores higher on coding benchmarks: 57.7 versus 44.0 in the Noometry coding category.
Which has the bigger context window?
GPT-5.6 Terra does, with 1.05M tokens against 400K.
How many benchmarks do GPT-5.6 Terra and GPT-5 Pro share?
8 benchmarks have published results for both models. GPT-5.6 Terra has 52 scored results on Noometry and GPT-5 Pro has 12.