Model comparison
Gemini 3 Pro vs GPT-5.6 Terra
GPT-5.6 Terra is the stronger model overall, scoring 59.2 to 54.8 on the Noometry Index.
Last verified . 36 shared benchmarks.
GoogleSummary
- They share 36 benchmarks with published results for both. Gemini 3 Pro scores higher in 4 categories and GPT-5.6 Terra in 6 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Terra leads 81.6 to 49.9.
- The biggest single-benchmark swing is ProofBench: 20% for Gemini 3 Pro and 74% for GPT-5.6 Terra.
Side by side
| Gemini 3 Pro | GPT-5.6 Terra | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 54.8 | 59.2 |
| Released | 2025-11-18 | 2026-07-09 |
| Weights | Proprietary | Proprietary |
| Context window | — | 1.05M |
| Max output | — | 128K |
| Input $ / M tokens | — | $2 |
| Output $ / M tokens | — | $12 |
| Results tracked | 67 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.6 Terra leads
Gemini 3 Pro: 51.6 (#39), GPT-5.6 Terra: 57.7 (#19)
| Benchmark | Gemini 3 Pro | GPT-5.6 Terra |
|---|---|---|
| LMArena WebDev | 1440 | 1522 |
| WeirdML | 69.9% | 78.3% |
| LMArena Coding | 1481 | 1484 |
| ALE-Bench | 1,177 | 1,951 |
| SWE-bench Verified | 72.9% | — |
| DeepSWE | — | 69.6% |
| FrontierCode | — | 41.3% |
| SWE-bench Verified (bash only) | 74.2% | — |
| CursorBench | — | 41.3% |
| SWE-bench Multilingual | 68.7% | — |
| SciCode | — | 55% |
| GSO | 18.6% | — |
| AlgoTune | 1.83 | — |
Agentic & Tool Use Too close to call
Gemini 3 Pro: 40.6 (#23), GPT-5.6 Terra: 40.1 (#25)
| Benchmark | Gemini 3 Pro | GPT-5.6 Terra |
|---|---|---|
| BALROG | 58.1% | 53.2% |
| Vending-Bench 2 | 5,478 | 7,343 |
| Terminal-Bench | 69.4% | — |
| APEX-Agents | — | 58.2% |
| Berkeley Function Calling Leaderboard | 72.5% | — |
| GDPval | 40.3% | — |
| Remote Labor Index | 1.3% | — |
| τ²-bench Airline | 80.5% | — |
| τ²-bench Banking | 18% | — |
| τ²-bench Retail | 75.9% | — |
| τ²-bench Telecom | 91% | — |
| DeepResearch Bench | 46.3% | — |
| GDP.pdf | — | 24.7% |
| LMArena Search | 1207 | — |
| METR Time Horizons | 71% | — |
Reasoning GPT-5.6 Terra leads
Gemini 3 Pro: 52.5 (#31), GPT-5.6 Terra: 60.7 (#21)
| Benchmark | Gemini 3 Pro | GPT-5.6 Terra |
|---|---|---|
| ARC-AGI-2 | 31.1% | 83.9% |
| SimpleBench | 76.4% | 48.9% |
| Kagi LLM Benchmark | 80.1% | 51.3% |
| NYT Connections (extended) | 94.4% | 78.4% |
| ARC-AGI-1 | 75% | 96.5% |
| CritPt | 6.9% | 30% |
| Chess Puzzles | 31% | 54% |
| LMArena Hard Prompts | 1480 | 1468 |
| Epoch Capabilities Index | 152.92 | 159.62 |
| EnigmaEval | 18.2% | — |
| Mystery Game Puzzles | — | 35% |
| DTBench | — | 93.3% |
| LMCA | — | 55% |
| Surface Evolver Bench | — | 83.8% |
| ForecastBench | 61.2 | — |
Math GPT-5.6 Terra leads
Gemini 3 Pro: 49.9 (#59), GPT-5.6 Terra: 81.6 (#12)
| Benchmark | Gemini 3 Pro | GPT-5.6 Terra |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 91.4% | 99.7% |
| ProofBench | 20% | 74% |
| LMArena Math | 1476 | 1466 |
| FrontierMath (Tiers 1-3) | — | 86% |
| FrontierMath Tier 4 | — | 70.7% |
| MathArena Final-Answer Competitions | 67% | — |
| Omni-MATH | 55.5% | — |
| FrontierMath (Feb 2025 set) | 37.6% | — |
| FrontierMath Tier 4 (v1) | 18.8% | — |
Knowledge Gemini 3 Pro leads
Gemini 3 Pro: 64.4 (#16), GPT-5.6 Terra: 61.2 (#30)
| Benchmark | Gemini 3 Pro | GPT-5.6 Terra |
|---|---|---|
| GPQA Diamond | 92.6% | 93.3% |
| LMArena Expert | 1475 | 1492 |
| Humanity's Last Exam | 37.5% | — |
| SimpleQA Verified | — | 43.2% |
| MMLU-Pro | 90.3% | — |
| Vectara Hallucination Rate | 13.6% | — |
| GPQA (HELM) | 80.3% | — |
Multimodal Gemini 3 Pro leads
Gemini 3 Pro: 57.6 (#2), GPT-5.6 Terra: 47.3 (#11)
| Benchmark | Gemini 3 Pro | GPT-5.6 Terra |
|---|---|---|
| LMArena Vision | 1305 | 1271 |
| LMArena Document | 1434 | 1472 |
| GeoBench | 84% | — |
| VPCT | 91% | — |
| Blueprint-Bench 2 | — | 30.8% |
| Furniture Assembly | — | 54.2% |
Multilingual Gemini 3 Pro leads
Gemini 3 Pro: 56.9 (#16), GPT-5.6 Terra: 54.4 (#44)
| Benchmark | Gemini 3 Pro | GPT-5.6 Terra |
|---|---|---|
| LMArena Non-English | 1474 | 1439 |
| LMArena Chinese | 1523 | 1513 |
| LMArena French | 1492 | 1471 |
| LMArena German | 1515 | 1460 |
| LMArena Japanese | 1510 | 1457 |
| LMArena Korean | 1448 | 1425 |
| LMArena Russian | 1493 | 1450 |
| LMArena Spanish | 1470 | 1448 |
Instruction Following Too close to call
Gemini 3 Pro: 76.3 (#45), GPT-5.6 Terra: 76.4 (#40)
| Benchmark | Gemini 3 Pro | GPT-5.6 Terra |
|---|---|---|
| LMArena Instruction Following | 1458 | 1454 |
| IFEval | 87.7% | — |
Long Context Too close to call
Gemini 3 Pro: 44.0 (#79), GPT-5.6 Terra: 44.4 (#68)
| Benchmark | Gemini 3 Pro | GPT-5.6 Terra |
|---|---|---|
| LMArena Longer Query | 1471 | 1451 |
| CL-bench | 15.8% | — |
Writing & Preference GPT-5.6 Terra leads
Gemini 3 Pro: 66.4 (#35), GPT-5.6 Terra: 70.2 (#23)
| Benchmark | Gemini 3 Pro | GPT-5.6 Terra |
|---|---|---|
| LMArena Text | 1479 | 1447 |
| LMArena Creative Writing | 1482 | 1410 |
| EQ-Bench Creative Writing | 1525 | 1855 |
| LMArena Multi-Turn | 1484 | 1449 |
| WildBench | 85.9% | — |
| EQ-Bench 4 | — | 1234 |
Frequently asked questions
Is Gemini 3 Pro better than GPT-5.6 Terra?
GPT-5.6 Terra is the stronger model overall, scoring 59.2 to 54.8 on the Noometry Index.
Is Gemini 3 Pro or GPT-5.6 Terra better for coding?
GPT-5.6 Terra scores higher on coding benchmarks: 57.7 versus 51.6 in the Noometry coding category.
How many benchmarks do Gemini 3 Pro and GPT-5.6 Terra share?
36 benchmarks have published results for both models. Gemini 3 Pro has 67 scored results on Noometry and GPT-5.6 Terra has 52.