# GLM-5.2 vs Inkling

> GLM-5.2 is the stronger model overall, scoring 50.9 to 44.1 on the Noometry Index.

- Canonical page: https://noometry.com/compare/glm-5-2-vs-inkling
- Last updated: 2026-10-10
- Shared benchmarks: 39

## Summary

- They share 39 benchmarks with published results for both. GLM-5.2 scores higher in 9 categories and Inkling in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where GLM-5.2 leads 55.4 to 31.2.
- The biggest single-benchmark swing is WeirdML: 70.1% for GLM-5.2 and 32.3% for Inkling.
- GLM-5.2 is cheaper at $1.40 / $4.40 per million input/output tokens, against $1.87 / $4.68 for Inkling.
- GLM-5.2 accepts more context: 1M tokens versus 66K.

## Snapshot

| | GLM-5.2 | Inkling |
|---|---|---|
| Provider | Z.ai (Zhipu) | Thinking Machines Lab |
| Noometry Index | 50.9 | 44.1 |
| Rank | 38 | 75 |
| Context | 1M | 66K |
| Input $/M | $1.40 | $1.87 |
| Output $/M | $4.40 | $4.68 |
| Weights | Open | Open |

## Coding

- GLM-5.2: 50.9 (#37)
- Inkling: 34.1 (#229)

| Benchmark | GLM-5.2 | Inkling |
|---|---|---|
| FrontierCode | 24.5% | 14% |
| LMArena WebDev | 1604 | 1412 |
| SciCode | 50.5% | 47% |
| WeirdML | 70.1% | 32.3% |
| LMArena Coding | 1484 | 1463 |
| ALE-Bench | 1,047 | 946 |
| SWE-bench Verified | 78.7% | — |
| DeepSWE | 43.8% | — |
| FrontierSWE | — | 4.1% |

## Agentic & Tool Use

- GLM-5.2: 31.6 (#63)
- Inkling: 29.5 (#76)

| Benchmark | GLM-5.2 | Inkling |
|---|---|---|
| τ²-bench Banking | 37.1% | 25% |
| APEX-Agents | — | 33.8% |
| PostTrainBench | 31.7% | — |
| GBAEval | 0% | — |
| Vending-Bench 2 | 8,314 | — |

## Reasoning

- GLM-5.2: 42.1 (#46)
- Inkling: 40.6 (#48)

| Benchmark | GLM-5.2 | Inkling |
|---|---|---|
| ARC-AGI-2 | 22.8% | 36.5% |
| SimpleBench | 58.8% | 50% |
| ARC-AGI-1 | 77% | 79.5% |
| CritPt | 20.9% | 5.4% |
| Chess Puzzles | 21% | 21% |
| LMArena Hard Prompts | 1480 | 1452 |
| DTBench | 93.6% | 87.5% |
| LMCA | 45.8% | 37.6% |
| Epoch Capabilities Index | 151.76 | 148.6 |
| Kagi LLM Benchmark | 62.6% | — |
| NYT Connections (extended) | 74.3% | — |
| EBR-Bench | 9.5% | — |
| Mystery Game Puzzles | 19% | — |
| Surface Evolver Bench | 55.6% | — |

## Math

- GLM-5.2: 55.4 (#35)
- Inkling: 31.2 (#198)

| Benchmark | GLM-5.2 | Inkling |
|---|---|---|
| FrontierMath (Tiers 1-3) | 59.2% | 33.3% |
| FrontierMath Tier 4 | 29.3% | 4.9% |
| OTIS Mock AIME 2024-2025 | 86.4% | 88.9% |
| ProofBench | 35% | 0% |
| LMArena Math | 1482 | 1482 |
| MathArena Final-Answer Competitions | 67.6% | — |

## Knowledge

- GLM-5.2: 57.1 (#36)
- Inkling: 55.2 (#44)

| Benchmark | GLM-5.2 | Inkling |
|---|---|---|
| GPQA Diamond | 91.9% | 88.3% |
| SimpleQA Verified | 34.2% | 40.3% |
| LMArena Expert | 1484 | 1468 |

## Multilingual

- GLM-5.2: 56.0 (#25)
- Inkling: 54.2 (#49)

| Benchmark | GLM-5.2 | Inkling |
|---|---|---|
| LMArena Non-English | 1459 | 1434 |
| LMArena Chinese | 1518 | 1495 |
| LMArena French | 1476 | 1453 |
| LMArena German | 1470 | 1456 |
| LMArena Japanese | 1451 | 1428 |
| LMArena Korean | 1445 | 1401 |
| LMArena Russian | 1467 | 1430 |
| LMArena Spanish | 1477 | 1442 |

## Instruction Following

- GLM-5.2: 77.1 (#32)
- Inkling: 75.3 (#65)

| Benchmark | GLM-5.2 | Inkling |
|---|---|---|
| LMArena Instruction Following | 1465 | 1427 |

## Long Context

- GLM-5.2: 45.3 (#43)
- Inkling: 43.9 (#79)

| Benchmark | GLM-5.2 | Inkling |
|---|---|---|
| LMArena Longer Query | 1479 | 1436 |

## Writing & Preference

- GLM-5.2: 70.7 (#20)
- Inkling: 65.5 (#48)

| Benchmark | GLM-5.2 | Inkling |
|---|---|---|
| LMArena Text | 1470 | 1443 |
| LMArena Creative Writing | 1465 | 1390 |
| EQ-Bench Creative Writing | 1757 | 1611 |
| EQ-Bench 4 | 1222 | 1226 |
| LMArena Multi-Turn | 1468 | 1436 |

## FAQ

### Is GLM-5.2 better than Inkling?

GLM-5.2 is the stronger model overall, scoring 50.9 to 44.1 on the Noometry Index.

### Which is cheaper, GLM-5.2 or Inkling?

GLM-5.2 is cheaper. It lists at $1.40 per million input tokens and $4.40 per million output tokens; Inkling lists at $1.87 and $4.68.

### Is GLM-5.2 or Inkling better for coding?

GLM-5.2 scores higher on coding benchmarks: 50.9 versus 34.1 in the Noometry coding category.

### Which has the bigger context window?

GLM-5.2 does, with 1M tokens against 66K.

### How many benchmarks do GLM-5.2 and Inkling share?

39 benchmarks have published results for both models. GLM-5.2 has 50 scored results on Noometry and Inkling has 41.