# Claude 3 Haiku vs Inkling-Small

> Inkling-Small is the stronger model overall, scoring 46.5 to 25.7 on the Noometry Index.

- Canonical page: https://noometry.com/compare/claude-3-haiku-vs-inkling-small
- Last updated: 2026-10-10
- Shared benchmarks: 22

## Summary

- They share 22 benchmarks with published results for both. Claude 3 Haiku scores higher in 0 categories and Inkling-Small in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Inkling-Small leads 44.9 to 9.2.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 1.8% for Claude 3 Haiku and 90% for Inkling-Small.
- Inkling-Small has downloadable open weights; the other is API-only.

## Snapshot

| | Claude 3 Haiku | Inkling-Small |
|---|---|---|
| Provider | Anthropic | Thinking Machines Lab |
| Noometry Index | 25.7 | 46.5 |
| Rank | 333 | 57 |
| Context | — | 524K |
| Input $/M | — | $0.45 |
| Output $/M | — | $1.20 |
| Weights | Proprietary | Open |

## Coding

- Claude 3 Haiku: 26.1 (#320)
- Inkling-Small: 43.3 (#80)

| Benchmark | Claude 3 Haiku | Inkling-Small |
|---|---|---|
| LMArena Coding | 1199 | 1451 |
| LMArena WebDev | — | 1408 |
| SciCode | — | 48.7% |
| WeirdML | 9.8% | — |
| BigCodeBench Instruct | 39.4% | — |
| BigCodeBench Complete | 50.1% | — |
| CadEval | 12% | — |
| HumanEval+ | 68.9% | — |
| MBPP+ | 68.8% | — |

## Reasoning

- Claude 3 Haiku: 15.7 (#303)
- Inkling-Small: 39.0 (#53)

| Benchmark | Claude 3 Haiku | Inkling-Small |
|---|---|---|
| LMArena Hard Prompts | 1174 | 1424 |
| Epoch Capabilities Index | 118.3 | 150.13 |
| ARC-AGI-2 | — | 40.1% |
| Kagi LLM Benchmark | 34.2% | — |
| ARC-AGI-1 | — | 84% |
| CritPt | — | 8.3% |
| Chess Puzzles | — | 18% |
| Mystery Game Puzzles | — | 6% |
| DTBench | 50.1% | — |
| LMCA | 8.8% | — |
| ForecastBench | 53.2 | — |
| WinoGrande | 74.2% | — |

## Math

- Claude 3 Haiku: 9.2 (#313)
- Inkling-Small: 44.9 (#65)

| Benchmark | Claude 3 Haiku | Inkling-Small |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.8% | 90% |
| LMArena Math | 1188 | 1453 |
| FrontierMath (Tiers 1-3) | — | 46.3% |
| FrontierMath Tier 4 | — | 17.1% |
| ProofBench | — | 6% |
| MATH Level 5 | 14.9% | — |

## Knowledge

- Claude 3 Haiku: 17.3 (#280)
- Inkling-Small: 48.3 (#71)

| Benchmark | Claude 3 Haiku | Inkling-Small |
|---|---|---|
| GPQA Diamond | 36.3% | 88.5% |
| LMArena Expert | 1148 | 1442 |
| SimpleQA Verified | — | 19.1% |
| Confabulations | 34.2% | — |
| MMLU | 73.8% | — |

## Multimodal

- Claude 3 Haiku: 23.1 (#123)
- Inkling-Small: 38.4 (#59)

| Benchmark | Claude 3 Haiku | Inkling-Small |
|---|---|---|
| LMArena Vision | 950 | 1235 |
| ScienceQA | 72% | — |

## Multilingual

- Claude 3 Haiku: 36.1 (#238)
- Inkling-Small: 51.7 (#102)

| Benchmark | Claude 3 Haiku | Inkling-Small |
|---|---|---|
| LMArena Non-English | 1178 | 1400 |
| LMArena Chinese | 1155 | 1458 |
| LMArena French | 1194 | 1438 |
| LMArena German | 1174 | 1412 |
| LMArena Japanese | 1102 | 1399 |
| LMArena Korean | 1108 | 1355 |
| LMArena Russian | 1204 | 1392 |
| LMArena Spanish | 1165 | 1426 |

## Instruction Following

- Claude 3 Haiku: 61.5 (#242)
- Inkling-Small: 73.9 (#112)

| Benchmark | Claude 3 Haiku | Inkling-Small |
|---|---|---|
| LMArena Instruction Following | 1174 | 1397 |

## Long Context

- Claude 3 Haiku: 36.1 (#232)
- Inkling-Small: 42.7 (#113)

| Benchmark | Claude 3 Haiku | Inkling-Small |
|---|---|---|
| LMArena Longer Query | 1191 | 1401 |

## Writing & Preference

- Claude 3 Haiku: 29.8 (#286)
- Inkling-Small: 59.8 (#102)

| Benchmark | Claude 3 Haiku | Inkling-Small |
|---|---|---|
| LMArena Text | 1195 | 1413 |
| LMArena Creative Writing | 1157 | 1330 |
| EQ-Bench Creative Writing | 717 | 1491 |
| LMArena Multi-Turn | 1190 | 1420 |

## FAQ

### Is Claude 3 Haiku better than Inkling-Small?

Inkling-Small is the stronger model overall, scoring 46.5 to 25.7 on the Noometry Index.

### Is Claude 3 Haiku or Inkling-Small better for coding?

Inkling-Small scores higher on coding benchmarks: 43.3 versus 26.1 in the Noometry coding category.

### How many benchmarks do Claude 3 Haiku and Inkling-Small share?

22 benchmarks have published results for both models. Claude 3 Haiku has 37 scored results on Noometry and Inkling-Small has 33.