Model comparison
DeepSeek-V3.2-Speciale vs Inkling-Small
Inkling-Small is the stronger model overall, scoring 46.5 to 39.7 on the Noometry Index.
Last verified . 1 shared benchmarks.
DeepSeekSummary
- They share 1 benchmark with published results for both. DeepSeek-V3.2-Speciale scores higher in 0 categories and Inkling-Small in 3 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Inkling-Small leads 59.6 to 46.0.
- Inkling-Small is cheaper at $0.45 / $1.20 per million input/output tokens, against $0.58 / $1.68 for DeepSeek-V3.2-Speciale.
- Inkling-Small accepts more context: 524K tokens versus 128K.
Side by side
| DeepSeek-V3.2-Speciale | Inkling-Small | |
|---|---|---|
| Provider | DeepSeek | Thinking Machines Lab |
| Noometry Index | 39.7 | 46.5 |
| Released | 2025-12-01 | 2026-07-15 |
| Weights | Open | Open |
| Context window | 128K | 524K |
| Max output | 128K | 1.05M |
| Input $ / M tokens | $0.58 | $0.45 |
| Output $ / M tokens | $1.68 | $1.20 |
| Results tracked | 3 | 33 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Inkling-Small leads
DeepSeek-V3.2-Speciale: 40.4 (#140), Inkling-Small: 43.6 (#85)
| Benchmark | DeepSeek-V3.2-Speciale | Inkling-Small |
|---|---|---|
| LMArena WebDev | — | 1409 |
| SciCode | — | 48.7% |
| WeirdML | 46.7% | — |
| LMArena Coding | — | 1451 |
Reasoning Inkling-Small leads
DeepSeek-V3.2-Speciale: 32.9 (#73), Inkling-Small: 38.6 (#63)
| Benchmark | DeepSeek-V3.2-Speciale | Inkling-Small |
|---|---|---|
| ARC-AGI-2 | — | 40.1% |
| SimpleBench | 52.6% | — |
| ARC-AGI-1 | — | 84% |
| CritPt | — | 8.3% |
| Chess Puzzles | — | 18% |
| LMArena Hard Prompts | — | 1423 |
| Mystery Game Puzzles | — | 6% |
| Epoch Capabilities Index | — | 150.15 |
Math Not comparable
DeepSeek-V3.2-Speciale: —, Inkling-Small: 45.1 (#77)
| Benchmark | DeepSeek-V3.2-Speciale | Inkling-Small |
|---|---|---|
| FrontierMath (Tiers 1-3) | — | 46.3% |
| FrontierMath Tier 4 | — | 17.1% |
| OTIS Mock AIME 2024-2025 | — | 90% |
| ProofBench | — | 6% |
| LMArena Math | — | 1459 |
Knowledge Not comparable
DeepSeek-V3.2-Speciale: —, Inkling-Small: 48.2 (#77)
| Benchmark | DeepSeek-V3.2-Speciale | Inkling-Small |
|---|---|---|
| GPQA Diamond | — | 88.5% |
| SimpleQA Verified | — | 19.1% |
| LMArena Expert | — | 1442 |
Multimodal Not comparable
DeepSeek-V3.2-Speciale: —, Inkling-Small: 39.1 (#62)
| Benchmark | DeepSeek-V3.2-Speciale | Inkling-Small |
|---|---|---|
| LMArena Vision | — | 1235 |
Multilingual Not comparable
DeepSeek-V3.2-Speciale: —, Inkling-Small: 51.7 (#104)
| Benchmark | DeepSeek-V3.2-Speciale | Inkling-Small |
|---|---|---|
| LMArena Non-English | — | 1402 |
| LMArena Chinese | — | 1465 |
| LMArena French | — | 1436 |
| LMArena German | — | 1405 |
| LMArena Japanese | — | 1405 |
| LMArena Korean | — | 1363 |
| LMArena Russian | — | 1391 |
| LMArena Spanish | — | 1428 |
Instruction Following Not comparable
DeepSeek-V3.2-Speciale: —, Inkling-Small: 73.8 (#114)
| Benchmark | DeepSeek-V3.2-Speciale | Inkling-Small |
|---|---|---|
| LMArena Instruction Following | — | 1399 |
Long Context Not comparable
DeepSeek-V3.2-Speciale: —, Inkling-Small: 42.7 (#118)
| Benchmark | DeepSeek-V3.2-Speciale | Inkling-Small |
|---|---|---|
| LMArena Longer Query | — | 1401 |
Writing & Preference Inkling-Small leads
DeepSeek-V3.2-Speciale: 46.0 (#222), Inkling-Small: 59.6 (#107)
| Benchmark | DeepSeek-V3.2-Speciale | Inkling-Small |
|---|---|---|
| EQ-Bench Creative Writing | 1276 | 1491 |
| LMArena Text | — | 1414 |
| LMArena Creative Writing | — | 1331 |
| LMArena Multi-Turn | — | 1418 |
Frequently asked questions
Is DeepSeek-V3.2-Speciale better than Inkling-Small?
Inkling-Small is the stronger model overall, scoring 46.5 to 39.7 on the Noometry Index.
Which is cheaper, DeepSeek-V3.2-Speciale or Inkling-Small?
Inkling-Small is cheaper. It lists at $0.45 per million input tokens and $1.20 per million output tokens; DeepSeek-V3.2-Speciale lists at $0.58 and $1.68.
Is DeepSeek-V3.2-Speciale or Inkling-Small better for coding?
Inkling-Small scores higher on coding benchmarks: 43.6 versus 40.4 in the Noometry coding category.
Which has the bigger context window?
Inkling-Small does, with 524K tokens against 128K.
How many benchmarks do DeepSeek-V3.2-Speciale and Inkling-Small share?
1 benchmark has published results for both models. DeepSeek-V3.2-Speciale has 3 scored results on Noometry and Inkling-Small has 33.