# Inkling-Small vs Llama-3.3-70B-Instruct

> Inkling-Small is the stronger model overall, scoring 46.5 to 30.6 on the Noometry Index. Llama-3.3-70B-Instruct costs 4.1× 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/inkling-small-vs-llama-3-3-70b-instruct
- Last updated: 2026-10-10
- Shared benchmarks: 22

## Summary

- They share 22 benchmarks with published results for both. Inkling-Small scores higher in 8 categories and Llama-3.3-70B-Instruct in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Inkling-Small leads 45.1 to 15.3.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 90% for Inkling-Small and 5.1% for Llama-3.3-70B-Instruct.
- Llama-3.3-70B-Instruct is cheaper at $0.10 / $0.32 per million input/output tokens, against $0.45 / $1.20 for Inkling-Small.
- Inkling-Small accepts more context: 524K tokens versus 128K.

## Snapshot

| | Inkling-Small | Llama-3.3-70B-Instruct |
|---|---|---|
| Provider | Thinking Machines Lab | Meta |
| Noometry Index | 46.5 | 30.6 |
| Rank | 63 | 291 |
| Context | 524K | 128K |
| Input $/M | $0.45 | $0.10 |
| Output $/M | $1.20 | $0.32 |
| Weights | Open | Open |

## Coding

- Inkling-Small: 43.6 (#85)
- Llama-3.3-70B-Instruct: 31.0 (#290)

| Benchmark | Inkling-Small | Llama-3.3-70B-Instruct |
|---|---|---|
| SciCode | 48.7% | 26% |
| LMArena Coding | 1451 | 1268 |
| LMArena WebDev | 1409 | — |
| WeirdML | — | 14.4% |
| BigCodeBench Instruct | — | 46.9% |
| LiveBench Coding | — | 36.6% |
| BigCodeBench Complete | — | 57.5% |

## Agentic & Tool Use

- Inkling-Small: —
- Llama-3.3-70B-Instruct: 25.8 (#105)

| Benchmark | Inkling-Small | Llama-3.3-70B-Instruct |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 31.9% |
| BALROG | — | 23% |

## Reasoning

- Inkling-Small: 38.6 (#63)
- Llama-3.3-70B-Instruct: 14.1 (#327)

| Benchmark | Inkling-Small | Llama-3.3-70B-Instruct |
|---|---|---|
| CritPt | 8.3% | 0% |
| LMArena Hard Prompts | 1423 | 1257 |
| Epoch Capabilities Index | 150.15 | 127.33 |
| ARC-AGI-2 | 40.1% | — |
| SimpleBench | — | 19.9% |
| ARC-AGI-1 | 84% | — |
| Chess Puzzles | 18% | — |
| LiveBench Reasoning | — | 50.8% |
| Mystery Game Puzzles | 6% | — |
| DTBench | — | 59.5% |
| LiveBench Data Analysis | — | 49.5% |
| LMCA | — | 17.5% |
| ForecastBench | — | 58.6 |
| LiveBench | — | 50.2% |

## Math

- Inkling-Small: 45.1 (#77)
- Llama-3.3-70B-Instruct: 15.3 (#298)

| Benchmark | Inkling-Small | Llama-3.3-70B-Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 90% | 5.1% |
| LMArena Math | 1459 | 1267 |
| FrontierMath (Tiers 1-3) | 46.3% | — |
| FrontierMath Tier 4 | 17.1% | — |
| ProofBench | 6% | — |
| LiveBench Math | — | 42.2% |
| MATH Level 5 | — | 41.6% |

## Knowledge

- Inkling-Small: 48.2 (#77)
- Llama-3.3-70B-Instruct: 30.6 (#226)

| Benchmark | Inkling-Small | Llama-3.3-70B-Instruct |
|---|---|---|
| GPQA Diamond | 88.5% | 47.4% |
| LMArena Expert | 1442 | 1225 |
| SimpleQA Verified | 19.1% | — |
| Confabulations | — | 22.8% |
| Vectara Hallucination Rate | — | 4.1% |
| MMLU | — | 86.3% |

## Multimodal

- Inkling-Small: 39.1 (#62)
- Llama-3.3-70B-Instruct: —

| Benchmark | Inkling-Small | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Vision | 1235 | — |

## Multilingual

- Inkling-Small: 51.7 (#104)
- Llama-3.3-70B-Instruct: 39.9 (#220)

| Benchmark | Inkling-Small | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Non-English | 1402 | 1236 |
| LMArena Chinese | 1465 | 1217 |
| LMArena French | 1436 | 1281 |
| LMArena German | 1405 | 1251 |
| LMArena Japanese | 1405 | 1150 |
| LMArena Korean | 1363 | 1143 |
| LMArena Russian | 1391 | 1252 |
| LMArena Spanish | 1428 | 1270 |

## Instruction Following

- Inkling-Small: 73.8 (#114)
- Llama-3.3-70B-Instruct: 71.1 (#157)

| Benchmark | Inkling-Small | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Instruction Following | 1399 | 1242 |
| LiveBench Instruction Following | — | 82.7% |

## Long Context

- Inkling-Small: 42.7 (#118)
- Llama-3.3-70B-Instruct: 26.4 (#295)

| Benchmark | Inkling-Small | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Longer Query | 1401 | 1256 |
| Fiction.LiveBench | — | 33.3% |

## Writing & Preference

- Inkling-Small: 59.6 (#107)
- Llama-3.3-70B-Instruct: 47.6 (#207)

| Benchmark | Inkling-Small | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Text | 1414 | 1274 |
| LMArena Creative Writing | 1331 | 1250 |
| LMArena Multi-Turn | 1418 | 1280 |
| EQ-Bench Creative Writing | 1491 | — |
| LiveBench Language | — | 39.2% |

## FAQ

### Is Inkling-Small better than Llama-3.3-70B-Instruct?

Inkling-Small is the stronger model overall, scoring 46.5 to 30.6 on the Noometry Index. Llama-3.3-70B-Instruct costs 4.1× less per token, which makes it the better buy when Inkling-Small's lead doesn't matter for your workload.

### Which is cheaper, Inkling-Small or Llama-3.3-70B-Instruct?

Llama-3.3-70B-Instruct is cheaper. It lists at $0.10 per million input tokens and $0.32 per million output tokens; Inkling-Small lists at $0.45 and $1.20.

### Is Inkling-Small or Llama-3.3-70B-Instruct better for coding?

Inkling-Small scores higher on coding benchmarks: 43.6 versus 31.0 in the Noometry coding category.

### Which has the bigger context window?

Inkling-Small does, with 524K tokens against 128K.

### How many benchmarks do Inkling-Small and Llama-3.3-70B-Instruct share?

22 benchmarks have published results for both models. Inkling-Small has 33 scored results on Noometry and Llama-3.3-70B-Instruct has 43.