# gpt-oss-20b vs Inkling-Small

> Inkling-Small is the stronger model overall, scoring 46.5 to 32.5 on the Noometry Index. gpt-oss-20b costs 18× less per token, which makes it the better buy when Inkling-Small's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gpt-oss-20b-vs-inkling-small
- Last updated: 2026-10-11
- Shared benchmarks: 23

## Summary

- They share 23 benchmarks with published results for both. gpt-oss-20b scores higher in 0 categories and Inkling-Small in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Inkling-Small leads 59.6 to 35.5.
- The biggest single-benchmark swing is GPQA Diamond: 60.8% for gpt-oss-20b and 88.5% for Inkling-Small.
- gpt-oss-20b is cheaper at $0.018 / $0.09 per million input/output tokens, against $0.45 / $1.20 for Inkling-Small.
- Inkling-Small accepts more context: 524K tokens versus 131K.

## Snapshot

| | gpt-oss-20b | Inkling-Small |
|---|---|---|
| Provider | OpenAI | Thinking Machines Lab |
| Noometry Index | 32.5 | 46.5 |
| Rank | 255 | 63 |
| Context | 131K | 524K |
| Input $/M | $0.018 | $0.45 |
| Output $/M | $0.09 | $1.20 |
| Weights | Open | Open |

## Coding

- gpt-oss-20b: 37.6 (#192)
- Inkling-Small: 43.6 (#85)

| Benchmark | gpt-oss-20b | Inkling-Small |
|---|---|---|
| SciCode | 34.4% | 48.7% |
| LMArena Coding | 1306 | 1451 |
| LMArena WebDev | — | 1409 |
| WeirdML | 40.9% | — |
| ALE-Bench | 566.05 | — |

## Agentic & Tool Use

- gpt-oss-20b: 9.3 (#154)
- Inkling-Small: —

| Benchmark | gpt-oss-20b | Inkling-Small |
|---|---|---|
| Terminal-Bench | 3.4% | — |

## Reasoning

- gpt-oss-20b: 19.3 (#261)
- Inkling-Small: 38.6 (#63)

| Benchmark | gpt-oss-20b | Inkling-Small |
|---|---|---|
| CritPt | 1.4% | 8.3% |
| Chess Puzzles | 4% | 18% |
| LMArena Hard Prompts | 1274 | 1423 |
| Epoch Capabilities Index | 137.82 | 150.15 |
| ARC-AGI-2 | — | 40.1% |
| Kagi LLM Benchmark | 53.2% | — |
| ARC-AGI-1 | — | 84% |
| Mystery Game Puzzles | — | 6% |
| DTBench | 68% | — |
| LMCA | 14.5% | — |

## Math

- gpt-oss-20b: 39.4 (#103)
- Inkling-Small: 45.1 (#77)

| Benchmark | gpt-oss-20b | Inkling-Small |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 65.3% | 90% |
| LMArena Math | 1317 | 1459 |
| FrontierMath (Tiers 1-3) | — | 46.3% |
| FrontierMath Tier 4 | — | 17.1% |
| ProofBench | — | 6% |
| Omni-MATH | 56.5% | — |

## Knowledge

- gpt-oss-20b: 34.6 (#195)
- Inkling-Small: 48.2 (#77)

| Benchmark | gpt-oss-20b | Inkling-Small |
|---|---|---|
| GPQA Diamond | 60.8% | 88.5% |
| LMArena Expert | 1258 | 1442 |
| SimpleQA Verified | — | 19.1% |
| MMLU-Pro | 74% | — |
| GPQA (HELM) | 59.4% | — |

## Multimodal

- gpt-oss-20b: —
- Inkling-Small: 39.1 (#62)

| Benchmark | gpt-oss-20b | Inkling-Small |
|---|---|---|
| LMArena Vision | — | 1235 |

## Multilingual

- gpt-oss-20b: 42.2 (#197)
- Inkling-Small: 51.7 (#104)

| Benchmark | gpt-oss-20b | Inkling-Small |
|---|---|---|
| LMArena Non-English | 1268 | 1402 |
| LMArena Chinese | 1314 | 1465 |
| LMArena German | 1255 | 1405 |
| LMArena Japanese | 1244 | 1405 |
| LMArena Korean | 1236 | 1363 |
| LMArena Russian | 1278 | 1391 |
| LMArena Spanish | 1267 | 1428 |
| LMArena French | — | 1436 |

## Instruction Following

- gpt-oss-20b: 61.8 (#240)
- Inkling-Small: 73.8 (#114)

| Benchmark | gpt-oss-20b | Inkling-Small |
|---|---|---|
| LMArena Instruction Following | 1236 | 1399 |
| IFEval | 73.2% | — |

## Long Context

- gpt-oss-20b: 37.9 (#209)
- Inkling-Small: 42.7 (#118)

| Benchmark | gpt-oss-20b | Inkling-Small |
|---|---|---|
| LMArena Longer Query | 1250 | 1401 |

## Writing & Preference

- gpt-oss-20b: 35.5 (#265)
- Inkling-Small: 59.6 (#107)

| Benchmark | gpt-oss-20b | Inkling-Small |
|---|---|---|
| LMArena Text | 1287 | 1414 |
| LMArena Creative Writing | 1201 | 1331 |
| EQ-Bench Creative Writing | 666 | 1491 |
| LMArena Multi-Turn | 1268 | 1418 |
| WildBench | 73.7% | — |

## FAQ

### Is gpt-oss-20b better than Inkling-Small?

Inkling-Small is the stronger model overall, scoring 46.5 to 32.5 on the Noometry Index. gpt-oss-20b costs 18× less per token, which makes it the better buy when Inkling-Small's lead doesn't matter for your workload.

### Which is cheaper, gpt-oss-20b or Inkling-Small?

gpt-oss-20b is cheaper. It lists at $0.018 per million input tokens and $0.09 per million output tokens; Inkling-Small lists at $0.45 and $1.20.

### Is gpt-oss-20b or Inkling-Small better for coding?

Inkling-Small scores higher on coding benchmarks: 43.6 versus 37.6 in the Noometry coding category.

### Which has the bigger context window?

Inkling-Small does, with 524K tokens against 131K.

### How many benchmarks do gpt-oss-20b and Inkling-Small share?

23 benchmarks have published results for both models. gpt-oss-20b has 34 scored results on Noometry and Inkling-Small has 33.