Model comparison
GPT-5.4 Pro vs GPT-5.6 Sol
GPT-5.6 Sol is the stronger model overall, scoring 64.4 to 58.8 on the Noometry Index.
Last verified . 13 shared benchmarks.
Summary
- They share 13 benchmarks with published results for both. GPT-5.4 Pro scores higher in 1 category and GPT-5.6 Sol in 3 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in coding, where GPT-5.6 Sol leads 61.1 to 42.6.
- The biggest single-benchmark swing is WeirdML: 57.4% for GPT-5.4 Pro and 89.4% for GPT-5.6 Sol.
- GPT-5.6 Sol is cheaper at $4 / $20 per million input/output tokens, against $30 / $180 for GPT-5.4 Pro.
Side by side
| GPT-5.4 Pro | GPT-5.6 Sol | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 58.8 | 64.4 |
| Released | 2026-03-05 | 2026-07-09 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 1.05M |
| Max output | 128K | 128K |
| Input $ / M tokens | $30 | $4 |
| Output $ / M tokens | $180 | $20 |
| Results tracked | 16 | 64 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.6 Sol leads
GPT-5.4 Pro: 42.6 (#89), GPT-5.6 Sol: 61.1 (#7)
| Benchmark | GPT-5.4 Pro | GPT-5.6 Sol |
|---|---|---|
| WeirdML | 57.4% | 89.4% |
| DeepSWE | — | 72.7% |
| FrontierCode | — | 47.5% |
| CursorBench | — | 41.7% |
| LMArena WebDev | — | 1619 |
| FrontierSWE | — | 32.2% |
| SciCode | — | 57.1% |
| LMArena Coding | — | 1493 |
| MirrorCode | — | 20% |
| ALE-Bench | — | 2,177 |
Agentic & Tool Use Not comparable
GPT-5.4 Pro: —, GPT-5.6 Sol: 50.2 (#7)
| Benchmark | GPT-5.4 Pro | GPT-5.6 Sol |
|---|---|---|
| APEX-Agents | — | 51.4% |
| OSWorld 2.0 | — | 27.3% |
| τ²-bench Banking | — | 46.9% |
| PostTrainBench | — | 36.2% |
| BALROG | — | 60% |
| GBAEval | — | 52.6% |
| GDP.pdf | — | 30.7% |
| LMArena Search | — | 1257 |
| Vending-Bench 2 | — | 9,619 |
Reasoning GPT-5.6 Sol leads
GPT-5.4 Pro: 71.0 (#10), GPT-5.6 Sol: 74.7 (#6)
| Benchmark | GPT-5.4 Pro | GPT-5.6 Sol |
|---|---|---|
| ARC-AGI-2 | 83.3% | 92.5% |
| SimpleBench | 74.1% | 71.7% |
| ARC-AGI-1 | 94.5% | 97.5% |
| CritPt | 30% | 32.3% |
| Chess Puzzles | 58.6% | 64% |
| EnigmaEval | 23.8% | 37.1% |
| Epoch Capabilities Index | 159.12 | 161.99 |
| Kagi LLM Benchmark | — | 67% |
| NYT Connections (extended) | — | 93.8% |
| EBR-Bench | — | 44.8% |
| LMArena Hard Prompts | — | 1481 |
| Mystery Game Puzzles | — | 58% |
| DTBench | — | 96% |
| LMCA | — | 59.2% |
| Surface Evolver Bench | — | 93.1% |
| Bench to the Future 3 | — | 0.14 |
Math GPT-5.6 Sol leads
GPT-5.4 Pro: 72.0 (#17), GPT-5.6 Sol: 85.5 (#7)
| Benchmark | GPT-5.4 Pro | GPT-5.6 Sol |
|---|---|---|
| FrontierMath (Tiers 1-3) | 82.5% | 89.1% |
| FrontierMath Tier 4 | 58.5% | 82.9% |
| OTIS Mock AIME 2024-2025 | — | 100% |
| ProofBench | — | 83% |
| LMArena Math | — | 1472 |
| FrontierMath (Feb 2025 set) | 50% | — |
| FrontierMath Erdős | — | 0% |
| FrontierMath Tier 4 (v1) | 37.5% | — |
Knowledge GPT-5.4 Pro leads
GPT-5.4 Pro: 68.3 (#6), GPT-5.6 Sol: 64.3 (#16)
| Benchmark | GPT-5.4 Pro | GPT-5.6 Sol |
|---|---|---|
| GPQA Diamond | 94.6% | 93.5% |
| SimpleQA Verified | 46.3% | 69.7% |
| Vectara Hallucination Rate | 8.3% | 12.4% |
| Humanity's Last Exam | 44.3% | — |
| LMArena Expert | — | 1517 |
Multimodal Not comparable
GPT-5.4 Pro: —, GPT-5.6 Sol: 48.9 (#8)
| Benchmark | GPT-5.4 Pro | GPT-5.6 Sol |
|---|---|---|
| LMArena Vision | — | 1281 |
| Blueprint-Bench 2 | — | 33.6% |
| Furniture Assembly | — | 56.7% |
| LMArena Document | — | 1483 |
Multilingual Not comparable
GPT-5.4 Pro: —, GPT-5.6 Sol: 55.4 (#32)
| Benchmark | GPT-5.4 Pro | GPT-5.6 Sol |
|---|---|---|
| LMArena Non-English | — | 1450 |
| LMArena Chinese | — | 1523 |
| LMArena French | — | 1475 |
| LMArena German | — | 1476 |
| LMArena Japanese | — | 1475 |
| LMArena Korean | — | 1434 |
| LMArena Russian | — | 1466 |
| LMArena Spanish | — | 1449 |
Instruction Following Not comparable
GPT-5.4 Pro: —, GPT-5.6 Sol: 77.8 (#16)
| Benchmark | GPT-5.4 Pro | GPT-5.6 Sol |
|---|---|---|
| LMArena Instruction Following | — | 1480 |
Long Context Not comparable
GPT-5.4 Pro: —, GPT-5.6 Sol: 45.3 (#42)
| Benchmark | GPT-5.4 Pro | GPT-5.6 Sol |
|---|---|---|
| LMArena Longer Query | — | 1479 |
Writing & Preference Not comparable
GPT-5.4 Pro: —, GPT-5.6 Sol: 73.5 (#12)
| Benchmark | GPT-5.4 Pro | GPT-5.6 Sol |
|---|---|---|
| LMArena Text | — | 1456 |
| LMArena Creative Writing | — | 1450 |
| EQ-Bench Creative Writing | — | 1972 |
| EQ-Bench 4 | — | 1250 |
| LMArena Multi-Turn | — | 1456 |
Frequently asked questions
Is GPT-5.4 Pro better than GPT-5.6 Sol?
GPT-5.6 Sol is the stronger model overall, scoring 64.4 to 58.8 on the Noometry Index.
Which is cheaper, GPT-5.4 Pro or GPT-5.6 Sol?
GPT-5.6 Sol is cheaper. It lists at $4 per million input tokens and $20 per million output tokens; GPT-5.4 Pro lists at $30 and $180.
Is GPT-5.4 Pro or GPT-5.6 Sol better for coding?
GPT-5.6 Sol scores higher on coding benchmarks: 61.1 versus 42.6 in the Noometry coding category.
Which has the bigger context window?
Both accept 1.05M tokens.
How many benchmarks do GPT-5.4 Pro and GPT-5.6 Sol share?
13 benchmarks have published results for both models. GPT-5.4 Pro has 16 scored results on Noometry and GPT-5.6 Sol has 64.