Model comparison
GPT-5.4 mini vs GPT-6 Sol
GPT-6 Sol is the stronger model overall, scoring 61.8 to 45.0 on the Noometry Index. GPT-5.4 mini costs 2.4× less per token, which makes it the better buy when GPT-6 Sol's lead doesn't matter for your workload.
Last verified . 38 shared benchmarks.
Summary
- They share 38 benchmarks with published results for both. GPT-5.4 mini scores higher in 1 category and GPT-6 Sol in 9 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-6 Sol leads 74.0 to 30.4.
- The biggest single-benchmark swing is FrontierMath Tier 4: 9.8% for GPT-5.4 mini and 90% for GPT-6 Sol.
- GPT-5.4 mini is cheaper at $0.75 / $4.50 per million input/output tokens, against $2 / $10 for GPT-6 Sol.
- GPT-6 Sol accepts more context: 1.05M tokens versus 400K.
Side by side
| GPT-5.4 mini | GPT-6 Sol | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 45.0 | 61.8 |
| Released | 2026-03-17 | 2026-09-22 |
| Weights | Proprietary | Proprietary |
| Context window | 400K | 1.05M |
| Max output | 128K | 128K |
| Input $ / M tokens | $0.75 | $2 |
| Output $ / M tokens | $4.50 | $10 |
| Results tracked | 46 | 45 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-6 Sol leads
GPT-5.4 mini: 45.2 (#72), GPT-6 Sol: 60.1 (#11)
| Benchmark | GPT-5.4 mini | GPT-6 Sol |
|---|---|---|
| FrontierCode | 27% | 49.3% |
| LMArena WebDev | 1397 | 1688 |
| SciCode | 49.9% | 57.6% |
| LMArena Coding | 1438 | 1447 |
| ALE-Bench | 1,189 | 2,462 |
| DeepSWE | — | 68.8% |
| WeirdML | 60.3% | — |
Agentic & Tool Use GPT-6 Sol leads
GPT-5.4 mini: 29.9 (#81), GPT-6 Sol: 37.2 (#36)
| Benchmark | GPT-5.4 mini | GPT-6 Sol |
|---|---|---|
| APEX-Agents | — | 54.3% |
| DeepResearch Bench | 36.3% | — |
| GDP.pdf | — | 26.4% |
| Vending-Bench 2 | — | 14,428 |
Reasoning GPT-6 Sol leads
GPT-5.4 mini: 30.4 (#85), GPT-6 Sol: 74.0 (#9)
| Benchmark | GPT-5.4 mini | GPT-6 Sol |
|---|---|---|
| ARC-AGI-2 | 18.9% | 89.6% |
| NYT Connections (extended) | 61.8% | 90.1% |
| ARC-AGI-1 | 63.7% | 95.5% |
| CritPt | 10% | 30.9% |
| LMArena Hard Prompts | 1424 | 1418 |
| Mystery Game Puzzles | 11% | 56% |
| DTBench | 80% | 97.3% |
| LMCA | 40.8% | 59.1% |
| Epoch Capabilities Index | 148.84 | 162.72 |
| Kagi LLM Benchmark | 37.9% | — |
| Chess Puzzles | 24% | — |
| Thematic Generalization | 61.7% | — |
| EBR-Bench | — | 53.3% |
| ForecastBench | 57 | — |
Math GPT-6 Sol leads
GPT-5.4 mini: 45.5 (#75), GPT-6 Sol: 87.2 (#7)
| Benchmark | GPT-5.4 mini | GPT-6 Sol |
|---|---|---|
| FrontierMath (Tiers 1-3) | 51.2% | 89.8% |
| FrontierMath Tier 4 | 9.8% | 90% |
| OTIS Mock AIME 2024-2025 | 88.9% | 100% |
| ProofBench | 21% | 83% |
| LMArena Math | 1419 | 1402 |
| FrontierMath (Feb 2025 set) | 28.3% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge GPT-6 Sol leads
GPT-5.4 mini: 51.5 (#67), GPT-6 Sol: 64.8 (#15)
| Benchmark | GPT-5.4 mini | GPT-6 Sol |
|---|---|---|
| GPQA Diamond | 86.9% | 94.3% |
| SimpleQA Verified | 29.4% | 60.7% |
| Vectara Hallucination Rate | 5.5% | 6.5% |
| LMArena Expert | 1435 | 1439 |
Multimodal GPT-6 Sol leads
GPT-5.4 mini: 39.7 (#56), GPT-6 Sol: 47.6 (#10)
| Benchmark | GPT-5.4 mini | GPT-6 Sol |
|---|---|---|
| LMArena Vision | 1245 | 1245 |
| Blueprint-Bench 2 | — | 36.9% |
| Furniture Assembly | — | 58.3% |
Multilingual GPT-5.4 mini leads
GPT-5.4 mini: 51.9 (#96), GPT-6 Sol: 50.5 (#118)
| Benchmark | GPT-5.4 mini | GPT-6 Sol |
|---|---|---|
| LMArena Non-English | 1405 | 1385 |
| LMArena Chinese | 1446 | 1405 |
| LMArena French | 1440 | 1410 |
| LMArena German | 1409 | 1390 |
| LMArena Japanese | 1374 | 1385 |
| LMArena Korean | 1368 | 1341 |
| LMArena Russian | 1417 | 1401 |
| LMArena Spanish | 1405 | 1384 |
Instruction Following Too close to call
GPT-5.4 mini: 74.1 (#102), GPT-6 Sol: 74.5 (#94)
| Benchmark | GPT-5.4 mini | GPT-6 Sol |
|---|---|---|
| LMArena Instruction Following | 1405 | 1412 |
Long Context Too close to call
GPT-5.4 mini: 43.0 (#112), GPT-6 Sol: 43.1 (#108)
| Benchmark | GPT-5.4 mini | GPT-6 Sol |
|---|---|---|
| LMArena Longer Query | 1407 | 1411 |
Writing & Preference GPT-6 Sol leads
GPT-5.4 mini: 64.0 (#58), GPT-6 Sol: 71.9 (#18)
| Benchmark | GPT-5.4 mini | GPT-6 Sol |
|---|---|---|
| LMArena Text | 1412 | 1395 |
| LMArena Creative Writing | 1370 | 1378 |
| EQ-Bench Creative Writing | 1665 | 2125 |
| LMArena Multi-Turn | 1429 | 1412 |
Frequently asked questions
Is GPT-5.4 mini better than GPT-6 Sol?
GPT-6 Sol is the stronger model overall, scoring 61.8 to 45.0 on the Noometry Index. GPT-5.4 mini costs 2.4× less per token, which makes it the better buy when GPT-6 Sol's lead doesn't matter for your workload.
Which is cheaper, GPT-5.4 mini or GPT-6 Sol?
GPT-5.4 mini is cheaper. It lists at $0.75 per million input tokens and $4.50 per million output tokens; GPT-6 Sol lists at $2 and $10.
Is GPT-5.4 mini or GPT-6 Sol better for coding?
GPT-6 Sol scores higher on coding benchmarks: 60.1 versus 45.2 in the Noometry coding category.
Which has the bigger context window?
GPT-6 Sol does, with 1.05M tokens against 400K.
How many benchmarks do GPT-5.4 mini and GPT-6 Sol share?
38 benchmarks have published results for both models. GPT-5.4 mini has 46 scored results on Noometry and GPT-6 Sol has 45.