# GPT-5.6 Terra vs Inkling

> GPT-5.6 Terra is the stronger model overall, scoring 59.1 to 44.1 on the Noometry Index. Inkling costs 1.7× less per token, which makes it the better buy when GPT-5.6 Terra's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gpt-5-6-terra-vs-inkling
- Last updated: 2026-10-10
- Shared benchmarks: 39

## Summary

- They share 39 benchmarks with published results for both. GPT-5.6 Terra scores higher in 9 categories and Inkling in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Terra leads 81.5 to 31.2.
- The biggest single-benchmark swing is ProofBench: 74% for GPT-5.6 Terra and 0% for Inkling.
- Inkling is cheaper at $1.87 / $4.68 per million input/output tokens, against $2 / $12 for GPT-5.6 Terra.
- GPT-5.6 Terra accepts more context: 1.05M tokens versus 66K.
- Inkling has downloadable open weights; the other is API-only.

## Snapshot

| | GPT-5.6 Terra | Inkling |
|---|---|---|
| Provider | OpenAI | Thinking Machines Lab |
| Noometry Index | 59.1 | 44.1 |
| Rank | 15 | 75 |
| Context | 1.05M | 66K |
| Input $/M | $2 | $1.87 |
| Output $/M | $12 | $4.68 |
| Weights | Proprietary | Open |

## Coding

- GPT-5.6 Terra: 57.1 (#16)
- Inkling: 34.1 (#229)

| Benchmark | GPT-5.6 Terra | Inkling |
|---|---|---|
| FrontierCode | 41.3% | 14% |
| LMArena WebDev | 1520 | 1412 |
| SciCode | 55% | 47% |
| WeirdML | 78.3% | 32.3% |
| LMArena Coding | 1484 | 1463 |
| ALE-Bench | 1,951 | 946 |
| DeepSWE | 69.6% | — |
| CursorBench | 41.3% | — |
| FrontierSWE | — | 4.1% |

## Agentic & Tool Use

- GPT-5.6 Terra: 40.2 (#23)
- Inkling: 29.5 (#76)

| Benchmark | GPT-5.6 Terra | Inkling |
|---|---|---|
| APEX-Agents | 58.2% | 33.8% |
| τ²-bench Banking | — | 25% |
| BALROG | 53.2% | — |
| GDP.pdf | 24.7% | — |
| Vending-Bench 2 | 7,343 | — |

## Reasoning

- GPT-5.6 Terra: 60.6 (#19)
- Inkling: 40.6 (#48)

| Benchmark | GPT-5.6 Terra | Inkling |
|---|---|---|
| ARC-AGI-2 | 83.9% | 36.5% |
| SimpleBench | 48.9% | 50% |
| ARC-AGI-1 | 96.5% | 79.5% |
| CritPt | 30% | 5.4% |
| Chess Puzzles | 54% | 21% |
| LMArena Hard Prompts | 1467 | 1452 |
| DTBench | 93.3% | 87.5% |
| LMCA | 55% | 37.6% |
| Epoch Capabilities Index | 159.31 | 148.6 |
| Kagi LLM Benchmark | 51.3% | — |
| NYT Connections (extended) | 78.4% | — |
| Mystery Game Puzzles | 35% | — |
| Surface Evolver Bench | 83.8% | — |

## Math

- GPT-5.6 Terra: 81.5 (#10)
- Inkling: 31.2 (#198)

| Benchmark | GPT-5.6 Terra | Inkling |
|---|---|---|
| FrontierMath (Tiers 1-3) | 86% | 33.3% |
| FrontierMath Tier 4 | 70.7% | 4.9% |
| OTIS Mock AIME 2024-2025 | 99.7% | 88.9% |
| ProofBench | 74% | 0% |
| LMArena Math | 1463 | 1482 |

## Knowledge

- GPT-5.6 Terra: 61.3 (#27)
- Inkling: 55.2 (#44)

| Benchmark | GPT-5.6 Terra | Inkling |
|---|---|---|
| GPQA Diamond | 93.3% | 88.3% |
| SimpleQA Verified | 43.2% | 40.3% |
| LMArena Expert | 1492 | 1468 |

## Multimodal

- GPT-5.6 Terra: 47.5 (#10)
- Inkling: —

| Benchmark | GPT-5.6 Terra | Inkling |
|---|---|---|
| LMArena Vision | 1271 | — |
| Blueprint-Bench 2 | 30.8% | — |
| Furniture Assembly | 54.2% | — |
| LMArena Document | 1472 | — |

## Multilingual

- GPT-5.6 Terra: 54.5 (#42)
- Inkling: 54.2 (#49)

| Benchmark | GPT-5.6 Terra | Inkling |
|---|---|---|
| LMArena Non-English | 1439 | 1434 |
| LMArena Chinese | 1514 | 1495 |
| LMArena French | 1476 | 1453 |
| LMArena German | 1462 | 1456 |
| LMArena Japanese | 1446 | 1428 |
| LMArena Korean | 1420 | 1401 |
| LMArena Russian | 1450 | 1430 |
| LMArena Spanish | 1455 | 1442 |

## Instruction Following

- GPT-5.6 Terra: 76.6 (#38)
- Inkling: 75.3 (#65)

| Benchmark | GPT-5.6 Terra | Inkling |
|---|---|---|
| LMArena Instruction Following | 1453 | 1427 |

## Long Context

- GPT-5.6 Terra: 44.4 (#64)
- Inkling: 43.9 (#79)

| Benchmark | GPT-5.6 Terra | Inkling |
|---|---|---|
| LMArena Longer Query | 1452 | 1436 |

## Writing & Preference

- GPT-5.6 Terra: 70.4 (#21)
- Inkling: 65.5 (#48)

| Benchmark | GPT-5.6 Terra | Inkling |
|---|---|---|
| LMArena Text | 1446 | 1443 |
| LMArena Creative Writing | 1411 | 1390 |
| EQ-Bench Creative Writing | 1855 | 1611 |
| EQ-Bench 4 | 1234 | 1226 |
| LMArena Multi-Turn | 1451 | 1436 |

## FAQ

### Is GPT-5.6 Terra better than Inkling?

GPT-5.6 Terra is the stronger model overall, scoring 59.1 to 44.1 on the Noometry Index. Inkling costs 1.7× less per token, which makes it the better buy when GPT-5.6 Terra's lead doesn't matter for your workload.

### Which is cheaper, GPT-5.6 Terra or Inkling?

Inkling is cheaper. It lists at $1.87 per million input tokens and $4.68 per million output tokens; GPT-5.6 Terra lists at $2 and $12.

### Is GPT-5.6 Terra or Inkling better for coding?

GPT-5.6 Terra scores higher on coding benchmarks: 57.1 versus 34.1 in the Noometry coding category.

### Which has the bigger context window?

GPT-5.6 Terra does, with 1.05M tokens against 66K.

### How many benchmarks do GPT-5.6 Terra and Inkling share?

39 benchmarks have published results for both models. GPT-5.6 Terra has 52 scored results on Noometry and Inkling has 41.