Model comparison
GPT-6 Sol vs Inkling-Small
GPT-6 Sol is the stronger model overall, scoring 61.8 to 46.5 on the Noometry Index. Inkling-Small costs 6.3× less per token, which makes it the better buy when GPT-6 Sol's lead doesn't matter for your workload.
Last verified . 32 shared benchmarks.
Summary
- They share 32 benchmarks with published results for both. GPT-6 Sol scores higher in 8 categories and Inkling-Small in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-6 Sol leads 87.2 to 45.1.
- The biggest single-benchmark swing is ProofBench: 83% for GPT-6 Sol and 6% for Inkling-Small.
- Inkling-Small is cheaper at $0.45 / $1.20 per million input/output tokens, against $2 / $10 for GPT-6 Sol.
- GPT-6 Sol accepts more context: 1.05M tokens versus 524K.
- Inkling-Small has downloadable open weights; the other is API-only.
Side by side
| GPT-6 Sol | Inkling-Small | |
|---|---|---|
| Provider | OpenAI | Thinking Machines Lab |
| Noometry Index | 61.8 | 46.5 |
| Released | 2026-09-22 | 2026-07-15 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 524K |
| Max output | 128K | 1.05M |
| Input $ / M tokens | $2 | $0.45 |
| Output $ / M tokens | $10 | $1.20 |
| Results tracked | 45 | 33 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-6 Sol leads
GPT-6 Sol: 60.1 (#11), Inkling-Small: 43.6 (#85)
| Benchmark | GPT-6 Sol | Inkling-Small |
|---|---|---|
| LMArena WebDev | 1688 | 1409 |
| SciCode | 57.6% | 48.7% |
| LMArena Coding | 1447 | 1451 |
| DeepSWE | 68.8% | — |
| FrontierCode | 49.3% | — |
| ALE-Bench | 2,462 | — |
Agentic & Tool Use Not comparable
GPT-6 Sol: 37.2 (#36), Inkling-Small: —
| Benchmark | GPT-6 Sol | Inkling-Small |
|---|---|---|
| APEX-Agents | 54.3% | — |
| GDP.pdf | 26.4% | — |
| Vending-Bench 2 | 14,428 | — |
Reasoning GPT-6 Sol leads
GPT-6 Sol: 74.0 (#9), Inkling-Small: 38.6 (#63)
| Benchmark | GPT-6 Sol | Inkling-Small |
|---|---|---|
| ARC-AGI-2 | 89.6% | 40.1% |
| ARC-AGI-1 | 95.5% | 84% |
| CritPt | 30.9% | 8.3% |
| LMArena Hard Prompts | 1418 | 1423 |
| Mystery Game Puzzles | 56% | 6% |
| Epoch Capabilities Index | 162.72 | 150.15 |
| NYT Connections (extended) | 90.1% | — |
| Chess Puzzles | — | 18% |
| EBR-Bench | 53.3% | — |
| DTBench | 97.3% | — |
| LMCA | 59.1% | — |
Math GPT-6 Sol leads
GPT-6 Sol: 87.2 (#7), Inkling-Small: 45.1 (#77)
| Benchmark | GPT-6 Sol | Inkling-Small |
|---|---|---|
| FrontierMath (Tiers 1-3) | 89.8% | 46.3% |
| FrontierMath Tier 4 | 90% | 17.1% |
| OTIS Mock AIME 2024-2025 | 100% | 90% |
| ProofBench | 83% | 6% |
| LMArena Math | 1402 | 1459 |
Knowledge GPT-6 Sol leads
GPT-6 Sol: 64.8 (#15), Inkling-Small: 48.2 (#77)
| Benchmark | GPT-6 Sol | Inkling-Small |
|---|---|---|
| GPQA Diamond | 94.3% | 88.5% |
| SimpleQA Verified | 60.7% | 19.1% |
| LMArena Expert | 1439 | 1442 |
| Vectara Hallucination Rate | 6.5% | — |
Multimodal GPT-6 Sol leads
GPT-6 Sol: 47.6 (#10), Inkling-Small: 39.1 (#62)
| Benchmark | GPT-6 Sol | Inkling-Small |
|---|---|---|
| LMArena Vision | 1245 | 1235 |
| Blueprint-Bench 2 | 36.9% | — |
| Furniture Assembly | 58.3% | — |
Multilingual Inkling-Small leads
GPT-6 Sol: 50.5 (#118), Inkling-Small: 51.7 (#104)
| Benchmark | GPT-6 Sol | Inkling-Small |
|---|---|---|
| LMArena Non-English | 1385 | 1402 |
| LMArena Chinese | 1405 | 1465 |
| LMArena French | 1410 | 1436 |
| LMArena German | 1390 | 1405 |
| LMArena Japanese | 1385 | 1405 |
| LMArena Korean | 1341 | 1363 |
| LMArena Russian | 1401 | 1391 |
| LMArena Spanish | 1384 | 1428 |
Instruction Following Too close to call
GPT-6 Sol: 74.5 (#94), Inkling-Small: 73.8 (#114)
| Benchmark | GPT-6 Sol | Inkling-Small |
|---|---|---|
| LMArena Instruction Following | 1412 | 1399 |
Long Context Too close to call
GPT-6 Sol: 43.1 (#108), Inkling-Small: 42.7 (#118)
| Benchmark | GPT-6 Sol | Inkling-Small |
|---|---|---|
| LMArena Longer Query | 1411 | 1401 |
Writing & Preference GPT-6 Sol leads
GPT-6 Sol: 71.9 (#18), Inkling-Small: 59.6 (#107)
| Benchmark | GPT-6 Sol | Inkling-Small |
|---|---|---|
| LMArena Text | 1395 | 1414 |
| LMArena Creative Writing | 1378 | 1331 |
| EQ-Bench Creative Writing | 2125 | 1491 |
| LMArena Multi-Turn | 1412 | 1418 |
Frequently asked questions
Is GPT-6 Sol better than Inkling-Small?
GPT-6 Sol is the stronger model overall, scoring 61.8 to 46.5 on the Noometry Index. Inkling-Small costs 6.3× 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-6 Sol or Inkling-Small?
Inkling-Small is cheaper. It lists at $0.45 per million input tokens and $1.20 per million output tokens; GPT-6 Sol lists at $2 and $10.
Is GPT-6 Sol or Inkling-Small better for coding?
GPT-6 Sol scores higher on coding benchmarks: 60.1 versus 43.6 in the Noometry coding category.
Which has the bigger context window?
GPT-6 Sol does, with 1.05M tokens against 524K.
How many benchmarks do GPT-6 Sol and Inkling-Small share?
32 benchmarks have published results for both models. GPT-6 Sol has 45 scored results on Noometry and Inkling-Small has 33.