# GPT-4 Turbo vs Inkling

> Inkling is the stronger model overall, scoring 44.1 to 30.3 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gpt-4-turbo-vs-inkling
- Last updated: 2026-10-10
- Shared benchmarks: 26

## Summary

- They share 26 benchmarks with published results for both. GPT-4 Turbo scores higher in 0 categories and Inkling in 8 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Inkling leads 55.2 to 24.3.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 6.7% for GPT-4 Turbo and 88.9% for Inkling.
- Inkling is cheaper at $1.87 / $4.68 per million input/output tokens, against $10 / $30 for GPT-4 Turbo.
- GPT-4 Turbo accepts more context: 128K tokens versus 66K.
- Inkling has downloadable open weights; the other is API-only.

## Snapshot

| | GPT-4 Turbo | Inkling |
|---|---|---|
| Provider | OpenAI | Thinking Machines Lab |
| Noometry Index | 30.3 | 44.1 |
| Rank | 281 | 75 |
| Context | 128K | 66K |
| Input $/M | $10 | $1.87 |
| Output $/M | $30 | $4.68 |
| Weights | Proprietary | Open |

## Coding

- GPT-4 Turbo: 33.4 (#247)
- Inkling: 34.1 (#229)

| Benchmark | GPT-4 Turbo | Inkling |
|---|---|---|
| WeirdML | 18% | 32.3% |
| LMArena Coding | 1268 | 1463 |
| FrontierCode | — | 14% |
| LMArena WebDev | — | 1412 |
| FrontierSWE | — | 4.1% |
| SciCode | — | 47% |
| BigCodeBench Instruct | 48.2% | — |
| BigCodeBench Complete | 58.2% | — |
| ALE-Bench | — | 946 |
| HumanEval+ | 86.6% | — |
| MBPP+ | 73.3% | — |

## Agentic & Tool Use

- GPT-4 Turbo: —
- Inkling: 29.5 (#76)

| Benchmark | GPT-4 Turbo | Inkling |
|---|---|---|
| APEX-Agents | — | 33.8% |
| τ²-bench Banking | — | 25% |
| METR Time Horizons | 36.7% | — |

## Reasoning

- GPT-4 Turbo: 14.9 (#311)
- Inkling: 40.6 (#48)

| Benchmark | GPT-4 Turbo | Inkling |
|---|---|---|
| SimpleBench | 25.1% | 50% |
| Chess Puzzles | 6% | 21% |
| LMArena Hard Prompts | 1251 | 1452 |
| DTBench | 61.6% | 87.5% |
| LMCA | 9.8% | 37.6% |
| Epoch Capabilities Index | 127.25 | 148.6 |
| ARC-AGI-2 | — | 36.5% |
| ARC-AGI-1 | — | 79.5% |
| CritPt | — | 5.4% |
| ForecastBench | 59.4 | — |

## Math

- GPT-4 Turbo: 8.6 (#316)
- Inkling: 31.2 (#198)

| Benchmark | GPT-4 Turbo | Inkling |
|---|---|---|
| FrontierMath (Tiers 1-3) | 0.7% | 33.3% |
| OTIS Mock AIME 2024-2025 | 6.7% | 88.9% |
| LMArena Math | 1272 | 1482 |
| FrontierMath Tier 4 | — | 4.9% |
| ProofBench | — | 0% |
| MATH Level 5 | 46.7% | — |

## Knowledge

- GPT-4 Turbo: 24.3 (#263)
- Inkling: 55.2 (#44)

| Benchmark | GPT-4 Turbo | Inkling |
|---|---|---|
| GPQA Diamond | 46.6% | 88.3% |
| LMArena Expert | 1223 | 1468 |
| SimpleQA Verified | — | 40.3% |
| Confabulations | 28.4% | — |
| MMLU | 81.3% | — |

## Multimodal

- GPT-4 Turbo: 30.0 (#104)
- Inkling: —

| Benchmark | GPT-4 Turbo | Inkling |
|---|---|---|
| LMArena Vision | 1090 | — |

## Multilingual

- GPT-4 Turbo: 40.7 (#211)
- Inkling: 54.2 (#49)

| Benchmark | GPT-4 Turbo | Inkling |
|---|---|---|
| LMArena Non-English | 1245 | 1434 |
| LMArena Chinese | 1242 | 1495 |
| LMArena French | 1275 | 1453 |
| LMArena German | 1259 | 1456 |
| LMArena Japanese | 1194 | 1428 |
| LMArena Korean | 1187 | 1401 |
| LMArena Russian | 1259 | 1430 |
| LMArena Spanish | 1260 | 1442 |

## Instruction Following

- GPT-4 Turbo: 66.0 (#211)
- Inkling: 75.3 (#65)

| Benchmark | GPT-4 Turbo | Inkling |
|---|---|---|
| LMArena Instruction Following | 1250 | 1427 |

## Long Context

- GPT-4 Turbo: 38.0 (#200)
- Inkling: 43.9 (#79)

| Benchmark | GPT-4 Turbo | Inkling |
|---|---|---|
| LMArena Longer Query | 1254 | 1436 |

## Writing & Preference

- GPT-4 Turbo: 47.9 (#201)
- Inkling: 65.5 (#48)

| Benchmark | GPT-4 Turbo | Inkling |
|---|---|---|
| LMArena Text | 1272 | 1443 |
| LMArena Creative Writing | 1269 | 1390 |
| LMArena Multi-Turn | 1267 | 1436 |
| EQ-Bench Creative Writing | — | 1611 |
| EQ-Bench 4 | — | 1226 |

## FAQ

### Is GPT-4 Turbo better than Inkling?

Inkling is the stronger model overall, scoring 44.1 to 30.3 on the Noometry Index.

### Which is cheaper, GPT-4 Turbo or Inkling?

Inkling is cheaper. It lists at $1.87 per million input tokens and $4.68 per million output tokens; GPT-4 Turbo lists at $10 and $30.

### Is GPT-4 Turbo or Inkling better for coding?

They score almost the same on coding (33.4 vs 34.1); test both on your own repository before choosing.

### Which has the bigger context window?

GPT-4 Turbo does, with 128K tokens against 66K.

### How many benchmarks do GPT-4 Turbo and Inkling share?

26 benchmarks have published results for both models. GPT-4 Turbo has 36 scored results on Noometry and Inkling has 41.