Model comparison
Inkling-Small vs Qwen3 32B
Inkling-Small is the stronger model overall, scoring 46.5 to 39.2 on the Noometry Index.
Last verified . 19 shared benchmarks.
Alibaba (Qwen)Summary
- They share 19 benchmarks with published results for both. Inkling-Small scores higher in 7 categories and Qwen3 32B in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Inkling-Small leads 38.6 to 20.2.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 90% for Inkling-Small and 66.9% for Qwen3 32B.
- Inkling-Small is cheaper at $0.45 / $1.20 per million input/output tokens, against $0.70 / $2.80 for Qwen3 32B.
- Inkling-Small accepts more context: 524K tokens versus 131K.
Side by side
| Inkling-Small | Qwen3 32B | |
|---|---|---|
| Provider | Thinking Machines Lab | Alibaba (Qwen) |
| Noometry Index | 46.5 | 39.2 |
| Released | 2026-07-15 | 2025-04 |
| Weights | Open | Open |
| Context window | 524K | 131K |
| Max output | 1.05M | 16K |
| Input $ / M tokens | $0.45 | $0.70 |
| Output $ / M tokens | $1.20 | $2.80 |
| Results tracked | 33 | 26 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Inkling-Small leads
Inkling-Small: 43.6 (#85), Qwen3 32B: 37.7 (#190)
| Benchmark | Inkling-Small | Qwen3 32B |
|---|---|---|
| SciCode | 48.7% | 35.4% |
| LMArena Coding | 1451 | 1358 |
| Aider Polyglot | — | 40% |
| LMArena WebDev | 1409 | — |
Agentic & Tool Use Not comparable
Inkling-Small: —, Qwen3 32B: 32.6 (#62)
| Benchmark | Inkling-Small | Qwen3 32B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 48.7% |
Reasoning Inkling-Small leads
Inkling-Small: 38.6 (#63), Qwen3 32B: 20.2 (#241)
| Benchmark | Inkling-Small | Qwen3 32B |
|---|---|---|
| CritPt | 8.3% | 0.3% |
| Chess Puzzles | 18% | 5% |
| LMArena Hard Prompts | 1423 | 1334 |
| Epoch Capabilities Index | 150.15 | 138.51 |
| ARC-AGI-2 | 40.1% | — |
| Kagi LLM Benchmark | — | 54.9% |
| ARC-AGI-1 | 84% | — |
| Mystery Game Puzzles | 6% | — |
| DTBench | — | 67.5% |
| LMCA | — | 17.3% |
Math Inkling-Small leads
Inkling-Small: 45.1 (#77), Qwen3 32B: 39.7 (#99)
| Benchmark | Inkling-Small | Qwen3 32B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 90% | 66.9% |
| LMArena Math | 1459 | 1399 |
| FrontierMath (Tiers 1-3) | 46.3% | — |
| FrontierMath Tier 4 | 17.1% | — |
| ProofBench | 6% | — |
Knowledge Inkling-Small leads
Inkling-Small: 48.2 (#77), Qwen3 32B: 40.0 (#125)
| Benchmark | Inkling-Small | Qwen3 32B |
|---|---|---|
| GPQA Diamond | 88.5% | 65.7% |
| LMArena Expert | 1442 | 1362 |
| SimpleQA Verified | 19.1% | — |
| Vectara Hallucination Rate | — | 5.9% |
Multimodal Not comparable
Inkling-Small: 39.1 (#62), Qwen3 32B: —
| Benchmark | Inkling-Small | Qwen3 32B |
|---|---|---|
| LMArena Vision | 1235 | — |
Multilingual Inkling-Small leads
Inkling-Small: 51.7 (#104), Qwen3 32B: 45.6 (#167)
| Benchmark | Inkling-Small | Qwen3 32B |
|---|---|---|
| LMArena Non-English | 1402 | 1317 |
| LMArena Chinese | 1465 | 1357 |
| LMArena German | 1405 | 1341 |
| LMArena Russian | 1391 | 1311 |
| LMArena French | 1436 | — |
| LMArena Japanese | 1405 | — |
| LMArena Korean | 1363 | — |
| LMArena Spanish | 1428 | — |
Instruction Following Inkling-Small leads
Inkling-Small: 73.8 (#114), Qwen3 32B: 68.9 (#179)
| Benchmark | Inkling-Small | Qwen3 32B |
|---|---|---|
| LMArena Instruction Following | 1399 | 1305 |
Long Context Qwen3 32B leads
Inkling-Small: 42.7 (#118), Qwen3 32B: 43.8 (#87)
| Benchmark | Inkling-Small | Qwen3 32B |
|---|---|---|
| LMArena Longer Query | 1401 | 1327 |
| Fiction.LiveBench | — | 74.2% |
Writing & Preference Inkling-Small leads
Inkling-Small: 59.6 (#107), Qwen3 32B: 52.9 (#163)
| Benchmark | Inkling-Small | Qwen3 32B |
|---|---|---|
| LMArena Text | 1414 | 1340 |
| LMArena Creative Writing | 1331 | 1297 |
| LMArena Multi-Turn | 1418 | 1331 |
| EQ-Bench Creative Writing | 1491 | — |
Frequently asked questions
Is Inkling-Small better than Qwen3 32B?
Inkling-Small is the stronger model overall, scoring 46.5 to 39.2 on the Noometry Index.
Which is cheaper, Inkling-Small or Qwen3 32B?
Inkling-Small is cheaper. It lists at $0.45 per million input tokens and $1.20 per million output tokens; Qwen3 32B lists at $0.70 and $2.80.
Is Inkling-Small or Qwen3 32B better for coding?
Inkling-Small scores higher on coding benchmarks: 43.6 versus 37.7 in the Noometry coding category.
Which has the bigger context window?
Inkling-Small does, with 524K tokens against 131K.
How many benchmarks do Inkling-Small and Qwen3 32B share?
19 benchmarks have published results for both models. Inkling-Small has 33 scored results on Noometry and Qwen3 32B has 26.