GPT-5.4 Pro vs GPT-5.6 Terra
GPT-5.4 Pro Verdict
- GPT-5.4 Pro and GPT-5.6 Terra (xhigh) are level on quality (64.2 vs 64.2).
- GPT-5.4 Pro is stronger in instruction following, knowledge, multimodal, reasoning.
- GPT-5.6 Terra (xhigh) is stronger in agents & tools, coding, composite, human preference, long context, maths.
- GPT-5.6 Terra (xhigh) is 15× cheaper ($4.50 vs $67.50 per 1M blended).
- GPT-5.6 Terra (xhigh) streams 93.1× faster (93 vs 1 tokens per second).
| Metric | GPT-5.4 Pro | GPT-5.6 Terra (xhigh) |
|---|---|---|
| BenchLeader Index | 64.2 | 64.2 |
| Instruction following score | 67.3 | 65.0 |
| Knowledge score | 73.2 | 60.9 |
| Multimodal score | 69.8 | 63.0 |
| Reasoning score | 75.1 | 58.2 |
| Agents & tools score | – | 69.6 |
| Coding score | – | 63.7 |
| Composite score | – | 77.1 |
| Human preference score | – | 66.3 |
| Long context score | – | 66.0 |
| Maths score | – | 71.0 |
| Blended price $/M | $67.50 | $4.50 |
| Output speed | 1 tok/s | 93 tok/s |
| Time to first answer | 6.7 s | 32.5 s |
| Context window | 1.1M | 1M |
| Humanity's Last Exam | 44.3% | – |
| SimpleBench | 74.1% | 48.9% |
| SciCode | – | 51.6% |
| ProofBench | – | 74.0% |
| Epoch Capabilities Index | 158.9 | – |
| LMArena Text | – | 1466 |
| LMArena Hard Prompts | – | 1491 |
| LMArena Coding | – | 1519 |
| LMArena WebDev | – | 1521 |
| LMArena Vision | – | 1269 |
| LMArena Agent | – | 1.5 |
| AA Intelligence Index | – | 38.2 |
| IFBench | – | 66.3% |
| AA-LCR | – | 79.0% |
| MMMU-Pro | – | 79.5% |
| AA-Omniscience | – | -3.0 |
| Terminal-Bench Hard | – | 62.9% |
| GPQA Diamond (AA) | – | 90.8% |
| Humanity's Last Exam (AA) | – | 41.9% |
| SciCode (AA) | – | 52.3% |
| τ²-Bench Telecom (AA) | – | 80.4% |
| LiveCodeBench | – | 85.9% |
| MMLU-Pro | – | 86.7% |
| IOI | – | 65.3% |
| CorpFin | – | 65.3% |
| TaxEval | – | 76.2% |
| GPQA Diamond (Vals) | – | 90.9% |
| MultiChallenge | 69.2% | – |
| VISTA | 53.9% | – |
| MultiNRC | 62.3% | – |
| TutorBench | 56.6% | – |
| ARC-AGI-1 | – | 94.0% |
| ARC-AGI-2 | – | 74.2% |
| ARC-AGI-3 | – | 0.7% |
Data as of 2026-09-09. Best configuration of each model; every score links to its source on the model pages.