Model comparison
GPT-5.6 Terra vs Grok 4.3
GPT-5.6 Terra is the stronger model overall, scoring 59.1 to 43.6 on the Noometry Index. Grok 4.3 costs 2.9× less per token, which makes it the better buy when GPT-5.6 Terra's lead doesn't matter for your workload.
Last verified . 38 shared benchmarks.
Summary
- They share 38 benchmarks with published results for both. GPT-5.6 Terra scores higher in 10 categories and Grok 4.3 in 0 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Terra leads 81.5 to 45.8.
- The biggest single-benchmark swing is ProofBench: 74% for GPT-5.6 Terra and 11% for Grok 4.3.
- Grok 4.3 is cheaper at $1.25 / $2.50 per million input/output tokens, against $2 / $12 for GPT-5.6 Terra.
- GPT-5.6 Terra accepts more context: 1.05M tokens versus 1M.
Side by side
| GPT-5.6 Terra | Grok 4.3 | |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 59.1 | 43.6 |
| Released | 2026-07-09 | 2026-04-17 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 1M |
| Max output | 128K | 30K |
| Input $ / M tokens | $2 | $1.25 |
| Output $ / M tokens | $12 | $2.50 |
| Results tracked | 52 | 40 |
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: 57.1 (#16), Grok 4.3: 41.3 (#117)
| Benchmark | GPT-5.6 Terra | Grok 4.3 |
|---|---|---|
| LMArena WebDev | 1520 | 1357 |
| SciCode | 55% | 47.3% |
| WeirdML | 78.3% | 49.9% |
| LMArena Coding | 1484 | 1415 |
| ALE-Bench | 1,951 | 944.17 |
| DeepSWE | 69.6% | — |
| FrontierCode | 41.3% | — |
| CursorBench | 41.3% | — |
Agentic & Tool Use GPT-5.6 Terra leads
GPT-5.6 Terra: 40.2 (#23), Grok 4.3: 27.6 (#93)
| Benchmark | GPT-5.6 Terra | Grok 4.3 |
|---|---|---|
| GDP.pdf | 24.7% | 8% |
| Vending-Bench 2 | 7,343 | 35.26 |
| APEX-Agents | 58.2% | — |
| BALROG | 53.2% | — |
| LMArena Search | — | 1165 |
Reasoning GPT-5.6 Terra leads
GPT-5.6 Terra: 60.6 (#19), Grok 4.3: 34.9 (#60)
| Benchmark | GPT-5.6 Terra | Grok 4.3 |
|---|---|---|
| NYT Connections (extended) | 78.4% | 55.2% |
| CritPt | 30% | 8% |
| Chess Puzzles | 54% | 25% |
| LMArena Hard Prompts | 1467 | 1397 |
| DTBench | 93.3% | 90.7% |
| LMCA | 55% | 38.3% |
| Epoch Capabilities Index | 159.31 | 149.13 |
| ARC-AGI-2 | 83.9% | — |
| SimpleBench | 48.9% | — |
| Kagi LLM Benchmark | 51.3% | — |
| ARC-AGI-1 | 96.5% | — |
| Mystery Game Puzzles | 35% | — |
| Surface Evolver Bench | 83.8% | — |
| ForecastBench | — | 60.3 |
Math GPT-5.6 Terra leads
GPT-5.6 Terra: 81.5 (#10), Grok 4.3: 45.8 (#60)
| Benchmark | GPT-5.6 Terra | Grok 4.3 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 86% | 42.8% |
| FrontierMath Tier 4 | 70.7% | 14.6% |
| OTIS Mock AIME 2024-2025 | 99.7% | 93.3% |
| ProofBench | 74% | 11% |
| LMArena Math | 1463 | 1388 |
Knowledge GPT-5.6 Terra leads
GPT-5.6 Terra: 61.3 (#27), Grok 4.3: 52.5 (#57)
| Benchmark | GPT-5.6 Terra | Grok 4.3 |
|---|---|---|
| GPQA Diamond | 93.3% | 88.8% |
| SimpleQA Verified | 43.2% | 33.2% |
| LMArena Expert | 1492 | 1385 |
Multimodal GPT-5.6 Terra leads
GPT-5.6 Terra: 47.5 (#10), Grok 4.3: 31.6 (#96)
| Benchmark | GPT-5.6 Terra | Grok 4.3 |
|---|---|---|
| LMArena Vision | 1271 | 1231 |
| Blueprint-Bench 2 | 30.8% | 0% |
| Furniture Assembly | 54.2% | — |
| LMArena Document | 1472 | — |
Multilingual GPT-5.6 Terra leads
GPT-5.6 Terra: 54.5 (#42), Grok 4.3: 50.7 (#114)
| Benchmark | GPT-5.6 Terra | Grok 4.3 |
|---|---|---|
| LMArena Non-English | 1439 | 1385 |
| LMArena Chinese | 1514 | 1423 |
| LMArena French | 1476 | 1412 |
| LMArena German | 1462 | 1395 |
| LMArena Japanese | 1446 | 1379 |
| LMArena Korean | 1420 | 1357 |
| LMArena Russian | 1450 | 1399 |
| LMArena Spanish | 1455 | 1402 |
Instruction Following GPT-5.6 Terra leads
GPT-5.6 Terra: 76.6 (#38), Grok 4.3: 72.3 (#135)
| Benchmark | GPT-5.6 Terra | Grok 4.3 |
|---|---|---|
| LMArena Instruction Following | 1453 | 1366 |
Long Context GPT-5.6 Terra leads
GPT-5.6 Terra: 44.4 (#64), Grok 4.3: 42.5 (#118)
| Benchmark | GPT-5.6 Terra | Grok 4.3 |
|---|---|---|
| LMArena Longer Query | 1452 | 1393 |
Writing & Preference GPT-5.6 Terra leads
GPT-5.6 Terra: 70.4 (#21), Grok 4.3: 58.6 (#112)
| Benchmark | GPT-5.6 Terra | Grok 4.3 |
|---|---|---|
| LMArena Text | 1446 | 1397 |
| LMArena Creative Writing | 1411 | 1380 |
| EQ-Bench 4 | 1234 | 1075 |
| LMArena Multi-Turn | 1451 | 1406 |
| EQ-Bench Creative Writing | 1855 | — |
Frequently asked questions
Is GPT-5.6 Terra better than Grok 4.3?
GPT-5.6 Terra is the stronger model overall, scoring 59.1 to 43.6 on the Noometry Index. Grok 4.3 costs 2.9× 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, GPT-5.6 Terra or Grok 4.3?
Grok 4.3 is cheaper. It lists at $1.25 per million input tokens and $2.50 per million output tokens; GPT-5.6 Terra lists at $2 and $12.
Is GPT-5.6 Terra or Grok 4.3 better for coding?
GPT-5.6 Terra scores higher on coding benchmarks: 57.1 versus 41.3 in the Noometry coding category.
Which has the bigger context window?
GPT-5.6 Terra does, with 1.05M tokens against 1M.
How many benchmarks do GPT-5.6 Terra and Grok 4.3 share?
38 benchmarks have published results for both models. GPT-5.6 Terra has 52 scored results on Noometry and Grok 4.3 has 40.