Model comparison
Inkling-Small vs Nemotron 3 Ultra
Inkling-Small is the stronger model overall, scoring 46.5 to 44.7 on the Noometry Index.
Last verified . 25 shared benchmarks.
NVIDIASummary
- They share 25 benchmarks with published results for both. Inkling-Small scores higher in 3 categories and Nemotron 3 Ultra in 5 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Inkling-Small leads 44.9 to 34.4.
- The biggest single-benchmark swing is Mystery Game Puzzles: 6% for Inkling-Small and 20% for Nemotron 3 Ultra.
- Inkling-Small is cheaper at $0.45 / $1.20 per million input/output tokens, against $0.60 / $2.40 for Nemotron 3 Ultra.
- Inkling-Small accepts more context: 524K tokens versus 262K.
Side by side
| Inkling-Small | Nemotron 3 Ultra | |
|---|---|---|
| Provider | Thinking Machines Lab | NVIDIA |
| Noometry Index | 46.5 | 44.7 |
| Released | 2026-07-15 | 2026-06-04 |
| Weights | Open | Open |
| Context window | 524K | 262K |
| Max output | 1.05M | 128K |
| Input $ / M tokens | $0.45 | $0.60 |
| Output $ / M tokens | $1.20 | $2.40 |
| Results tracked | 33 | 29 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Inkling-Small leads
Inkling-Small: 43.3 (#80), Nemotron 3 Ultra: 37.7 (#174)
| Benchmark | Inkling-Small | Nemotron 3 Ultra |
|---|---|---|
| SciCode | 48.7% | 40.3% |
| LMArena Coding | 1451 | 1468 |
| FrontierCode | — | 13.6% |
| LMArena WebDev | 1408 | — |
| WeirdML | — | 43.5% |
Reasoning Inkling-Small leads
Inkling-Small: 39.0 (#53), Nemotron 3 Ultra: 30.3 (#78)
| Benchmark | Inkling-Small | Nemotron 3 Ultra |
|---|---|---|
| CritPt | 8.3% | 3.1% |
| Chess Puzzles | 18% | 12% |
| LMArena Hard Prompts | 1424 | 1452 |
| Mystery Game Puzzles | 6% | 20% |
| Epoch Capabilities Index | 150.13 | 146.2 |
| ARC-AGI-2 | 40.1% | — |
| ARC-AGI-1 | 84% | — |
| DTBench | — | 90.1% |
| LMCA | — | 36.9% |
Math Inkling-Small leads
Inkling-Small: 44.9 (#65), Nemotron 3 Ultra: 34.4 (#160)
| Benchmark | Inkling-Small | Nemotron 3 Ultra |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 90% | 86.7% |
| ProofBench | 6% | 2% |
| LMArena Math | 1453 | 1456 |
| FrontierMath (Tiers 1-3) | 46.3% | — |
| FrontierMath Tier 4 | 17.1% | — |
Knowledge Nemotron 3 Ultra leads
Inkling-Small: 48.3 (#71), Nemotron 3 Ultra: 52.5 (#56)
| Benchmark | Inkling-Small | Nemotron 3 Ultra |
|---|---|---|
| GPQA Diamond | 88.5% | 85.4% |
| LMArena Expert | 1442 | 1472 |
| SimpleQA Verified | 19.1% | — |
Multimodal Not comparable
Inkling-Small: 38.4 (#59), Nemotron 3 Ultra: —
| Benchmark | Inkling-Small | Nemotron 3 Ultra |
|---|---|---|
| LMArena Vision | 1235 | — |
Multilingual Nemotron 3 Ultra leads
Inkling-Small: 51.7 (#102), Nemotron 3 Ultra: 53.7 (#61)
| Benchmark | Inkling-Small | Nemotron 3 Ultra |
|---|---|---|
| LMArena Non-English | 1400 | 1427 |
| LMArena Chinese | 1458 | 1498 |
| LMArena French | 1438 | 1475 |
| LMArena German | 1412 | 1471 |
| LMArena Korean | 1355 | 1387 |
| LMArena Russian | 1392 | 1415 |
| LMArena Spanish | 1426 | 1452 |
| LMArena Japanese | 1399 | — |
Instruction Following Too close to call
Inkling-Small: 73.9 (#112), Nemotron 3 Ultra: 74.9 (#84)
| Benchmark | Inkling-Small | Nemotron 3 Ultra |
|---|---|---|
| LMArena Instruction Following | 1397 | 1417 |
Long Context Nemotron 3 Ultra leads
Inkling-Small: 42.7 (#113), Nemotron 3 Ultra: 43.9 (#77)
| Benchmark | Inkling-Small | Nemotron 3 Ultra |
|---|---|---|
| LMArena Longer Query | 1401 | 1437 |
Writing & Preference Nemotron 3 Ultra leads
Inkling-Small: 59.8 (#102), Nemotron 3 Ultra: 66.3 (#35)
| Benchmark | Inkling-Small | Nemotron 3 Ultra |
|---|---|---|
| LMArena Text | 1413 | 1444 |
| LMArena Creative Writing | 1330 | 1392 |
| EQ-Bench Creative Writing | 1491 | 1692 |
| LMArena Multi-Turn | 1420 | 1412 |
Frequently asked questions
Is Inkling-Small better than Nemotron 3 Ultra?
Inkling-Small is the stronger model overall, scoring 46.5 to 44.7 on the Noometry Index.
Which is cheaper, Inkling-Small or Nemotron 3 Ultra?
Inkling-Small is cheaper. It lists at $0.45 per million input tokens and $1.20 per million output tokens; Nemotron 3 Ultra lists at $0.60 and $2.40.
Is Inkling-Small or Nemotron 3 Ultra better for coding?
Inkling-Small scores higher on coding benchmarks: 43.3 versus 37.7 in the Noometry coding category.
Which has the bigger context window?
Inkling-Small does, with 524K tokens against 262K.
How many benchmarks do Inkling-Small and Nemotron 3 Ultra share?
25 benchmarks have published results for both models. Inkling-Small has 33 scored results on Noometry and Nemotron 3 Ultra has 29.