Model comparison
GPT-5.6 Sol vs Inkling
GPT-5.6 Sol is the stronger model overall, scoring 64.4 to 44.1 on the Noometry Index. Inkling costs 3.1× less per token, which makes it the better buy when GPT-5.6 Sol's lead doesn't matter for your workload.
Last verified . 41 shared benchmarks.
Summary
- They share 41 benchmarks with published results for both. GPT-5.6 Sol scores higher in 9 categories and Inkling in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Sol leads 85.5 to 31.2.
- The biggest single-benchmark swing is ProofBench: 83% for GPT-5.6 Sol and 0% for Inkling.
- Inkling is cheaper at $1.87 / $4.68 per million input/output tokens, against $4 / $20 for GPT-5.6 Sol.
- GPT-5.6 Sol accepts more context: 1.05M tokens versus 66K.
- Inkling has downloadable open weights; the other is API-only.
Side by side
| GPT-5.6 Sol | Inkling | |
|---|---|---|
| Provider | OpenAI | Thinking Machines Lab |
| Noometry Index | 64.4 | 44.1 |
| Released | 2026-07-09 | 2026-07-15 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 66K |
| Max output | 128K | 66K |
| Input $ / M tokens | $4 | $1.87 |
| Output $ / M tokens | $20 | $4.68 |
| Results tracked | 64 | 41 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.6 Sol leads
GPT-5.6 Sol: 61.1 (#7), Inkling: 34.1 (#229)
| Benchmark | GPT-5.6 Sol | Inkling |
|---|---|---|
| FrontierCode | 47.5% | 14% |
| LMArena WebDev | 1619 | 1412 |
| FrontierSWE | 32.2% | 4.1% |
| SciCode | 57.1% | 47% |
| WeirdML | 89.4% | 32.3% |
| LMArena Coding | 1493 | 1463 |
| ALE-Bench | 2,177 | 946 |
| DeepSWE | 72.7% | — |
| CursorBench | 41.7% | — |
| MirrorCode | 20% | — |
Agentic & Tool Use GPT-5.6 Sol leads
GPT-5.6 Sol: 50.2 (#7), Inkling: 29.5 (#76)
| Benchmark | GPT-5.6 Sol | Inkling |
|---|---|---|
| APEX-Agents | 51.4% | 33.8% |
| τ²-bench Banking | 46.9% | 25% |
| OSWorld 2.0 | 27.3% | — |
| 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-5.6 Sol: 74.7 (#6), Inkling: 40.6 (#48)
| Benchmark | GPT-5.6 Sol | Inkling |
|---|---|---|
| ARC-AGI-2 | 92.5% | 36.5% |
| SimpleBench | 71.7% | 50% |
| ARC-AGI-1 | 97.5% | 79.5% |
| CritPt | 32.3% | 5.4% |
| Chess Puzzles | 64% | 21% |
| LMArena Hard Prompts | 1481 | 1452 |
| DTBench | 96% | 87.5% |
| LMCA | 59.2% | 37.6% |
| Epoch Capabilities Index | 161.99 | 148.6 |
| Kagi LLM Benchmark | 67% | — |
| NYT Connections (extended) | 93.8% | — |
| EnigmaEval | 37.1% | — |
| EBR-Bench | 44.8% | — |
| Mystery Game Puzzles | 58% | — |
| Surface Evolver Bench | 93.1% | — |
| Bench to the Future 3 | 0.14 | — |
Math GPT-5.6 Sol leads
GPT-5.6 Sol: 85.5 (#7), Inkling: 31.2 (#198)
| Benchmark | GPT-5.6 Sol | Inkling |
|---|---|---|
| FrontierMath (Tiers 1-3) | 89.1% | 33.3% |
| FrontierMath Tier 4 | 82.9% | 4.9% |
| OTIS Mock AIME 2024-2025 | 100% | 88.9% |
| ProofBench | 83% | 0% |
| LMArena Math | 1472 | 1482 |
| FrontierMath Erdős | 0% | — |
Knowledge GPT-5.6 Sol leads
GPT-5.6 Sol: 64.3 (#16), Inkling: 55.2 (#44)
| Benchmark | GPT-5.6 Sol | Inkling |
|---|---|---|
| GPQA Diamond | 93.5% | 88.3% |
| SimpleQA Verified | 69.7% | 40.3% |
| LMArena Expert | 1517 | 1468 |
| Vectara Hallucination Rate | 12.4% | — |
Multimodal Not comparable
GPT-5.6 Sol: 48.9 (#8), Inkling: —
| Benchmark | GPT-5.6 Sol | Inkling |
|---|---|---|
| LMArena Vision | 1281 | — |
| Blueprint-Bench 2 | 33.6% | — |
| Furniture Assembly | 56.7% | — |
| LMArena Document | 1483 | — |
Multilingual GPT-5.6 Sol leads
GPT-5.6 Sol: 55.4 (#32), Inkling: 54.2 (#49)
| Benchmark | GPT-5.6 Sol | Inkling |
|---|---|---|
| LMArena Non-English | 1450 | 1434 |
| LMArena Chinese | 1523 | 1495 |
| LMArena French | 1475 | 1453 |
| LMArena German | 1476 | 1456 |
| LMArena Japanese | 1475 | 1428 |
| LMArena Korean | 1434 | 1401 |
| LMArena Russian | 1466 | 1430 |
| LMArena Spanish | 1449 | 1442 |
Instruction Following GPT-5.6 Sol leads
GPT-5.6 Sol: 77.8 (#16), Inkling: 75.3 (#65)
| Benchmark | GPT-5.6 Sol | Inkling |
|---|---|---|
| LMArena Instruction Following | 1480 | 1427 |
Long Context GPT-5.6 Sol leads
GPT-5.6 Sol: 45.3 (#42), Inkling: 43.9 (#79)
| Benchmark | GPT-5.6 Sol | Inkling |
|---|---|---|
| LMArena Longer Query | 1479 | 1436 |
Writing & Preference GPT-5.6 Sol leads
GPT-5.6 Sol: 73.5 (#12), Inkling: 65.5 (#48)
| Benchmark | GPT-5.6 Sol | Inkling |
|---|---|---|
| LMArena Text | 1456 | 1443 |
| LMArena Creative Writing | 1450 | 1390 |
| EQ-Bench Creative Writing | 1972 | 1611 |
| EQ-Bench 4 | 1250 | 1226 |
| LMArena Multi-Turn | 1456 | 1436 |
Frequently asked questions
Is GPT-5.6 Sol better than Inkling?
GPT-5.6 Sol is the stronger model overall, scoring 64.4 to 44.1 on the Noometry Index. Inkling costs 3.1× less per token, which makes it the better buy when GPT-5.6 Sol's lead doesn't matter for your workload.
Which is cheaper, GPT-5.6 Sol or Inkling?
Inkling is cheaper. It lists at $1.87 per million input tokens and $4.68 per million output tokens; GPT-5.6 Sol lists at $4 and $20.
Is GPT-5.6 Sol or Inkling better for coding?
GPT-5.6 Sol scores higher on coding benchmarks: 61.1 versus 34.1 in the Noometry coding category.
Which has the bigger context window?
GPT-5.6 Sol does, with 1.05M tokens against 66K.
How many benchmarks do GPT-5.6 Sol and Inkling share?
41 benchmarks have published results for both models. GPT-5.6 Sol has 64 scored results on Noometry and Inkling has 41.