# GLM-5.3-Flash vs Inkling-Small

> GLM-5.3-Flash is the stronger model overall, scoring 49.6 to 46.5 on the Noometry Index.

- Canonical page: https://noometry.com/compare/glm-5-3-flash-vs-inkling-small
- Last updated: 2026-10-10
- Shared benchmarks: 29

## Summary

- They share 29 benchmarks with published results for both. GLM-5.3-Flash scores higher in 8 categories and Inkling-Small in 1 category; 9 gaps are clear of the uncertainty.
- The widest gap is in coding, where GLM-5.3-Flash leads 53.7 to 43.3.
- The biggest single-benchmark swing is ProofBench: 21% for GLM-5.3-Flash and 6% for Inkling-Small.
- GLM-5.3-Flash is cheaper at $0.15 / $0.50 per million input/output tokens, against $0.45 / $1.20 for Inkling-Small.
- GLM-5.3-Flash accepts more context: 1M tokens versus 524K.

## Snapshot

| | GLM-5.3-Flash | Inkling-Small |
|---|---|---|
| Provider | Z.ai (Zhipu) | Thinking Machines Lab |
| Noometry Index | 49.6 | 46.5 |
| Rank | 45 | 57 |
| Context | 1M | 524K |
| Input $/M | $0.15 | $0.45 |
| Output $/M | $0.50 | $1.20 |
| Weights | Open | Open |

## Coding

- GLM-5.3-Flash: 53.7 (#26)
- Inkling-Small: 43.3 (#80)

| Benchmark | GLM-5.3-Flash | Inkling-Small |
|---|---|---|
| LMArena WebDev | 1614 | 1408 |
| SciCode | 51.6% | 48.7% |
| LMArena Coding | 1506 | 1451 |
| DeepSWE | 63.4% | — |
| FrontierCode | 31.8% | — |
| ALE-Bench | 303.55 | — |

## Agentic & Tool Use

- GLM-5.3-Flash: 34.3 (#41)
- Inkling-Small: —

| Benchmark | GLM-5.3-Flash | Inkling-Small |
|---|---|---|
| APEX-Agents | 52.8% | — |
| GDP.pdf | 14% | — |

## Reasoning

- GLM-5.3-Flash: 31.7 (#69)
- Inkling-Small: 39.0 (#53)

| Benchmark | GLM-5.3-Flash | Inkling-Small |
|---|---|---|
| CritPt | 15.4% | 8.3% |
| Chess Puzzles | 14% | 18% |
| LMArena Hard Prompts | 1490 | 1424 |
| Mystery Game Puzzles | 8% | 6% |
| Epoch Capabilities Index | 151.9 | 150.13 |
| ARC-AGI-2 | — | 40.1% |
| ARC-AGI-1 | — | 84% |
| Surface Evolver Bench | 52.5% | — |
| Bench to the Future 3 | 0.15 | — |

## Math

- GLM-5.3-Flash: 53.1 (#38)
- Inkling-Small: 44.9 (#65)

| Benchmark | GLM-5.3-Flash | Inkling-Small |
|---|---|---|
| FrontierMath (Tiers 1-3) | 55.8% | 46.3% |
| FrontierMath Tier 4 | 17.1% | 17.1% |
| OTIS Mock AIME 2024-2025 | 93.9% | 90% |
| ProofBench | 21% | 6% |
| LMArena Math | 1499 | 1453 |

## Knowledge

- GLM-5.3-Flash: 58.4 (#33)
- Inkling-Small: 48.3 (#71)

| Benchmark | GLM-5.3-Flash | Inkling-Small |
|---|---|---|
| GPQA Diamond | 90.2% | 88.5% |
| LMArena Expert | 1512 | 1442 |
| SimpleQA Verified | — | 19.1% |

## Multimodal

- GLM-5.3-Flash: 42.6 (#25)
- Inkling-Small: 38.4 (#59)

| Benchmark | GLM-5.3-Flash | Inkling-Small |
|---|---|---|
| LMArena Vision | 1304 | 1235 |

## Multilingual

- GLM-5.3-Flash: 56.2 (#24)
- Inkling-Small: 51.7 (#102)

| Benchmark | GLM-5.3-Flash | Inkling-Small |
|---|---|---|
| LMArena Non-English | 1462 | 1400 |
| LMArena Chinese | 1529 | 1458 |
| LMArena French | 1495 | 1438 |
| LMArena German | 1457 | 1412 |
| LMArena Japanese | 1420 | 1399 |
| LMArena Korean | 1443 | 1355 |
| LMArena Russian | 1472 | 1392 |
| LMArena Spanish | 1471 | 1426 |

## Instruction Following

- GLM-5.3-Flash: 77.8 (#17)
- Inkling-Small: 73.9 (#112)

| Benchmark | GLM-5.3-Flash | Inkling-Small |
|---|---|---|
| LMArena Instruction Following | 1480 | 1397 |

## Long Context

- GLM-5.3-Flash: 45.3 (#41)
- Inkling-Small: 42.7 (#113)

| Benchmark | GLM-5.3-Flash | Inkling-Small |
|---|---|---|
| LMArena Longer Query | 1479 | 1401 |

## Writing & Preference

- GLM-5.3-Flash: 65.4 (#49)
- Inkling-Small: 59.8 (#102)

| Benchmark | GLM-5.3-Flash | Inkling-Small |
|---|---|---|
| LMArena Text | 1471 | 1413 |
| LMArena Creative Writing | 1440 | 1330 |
| LMArena Multi-Turn | 1471 | 1420 |
| EQ-Bench Creative Writing | — | 1491 |

## FAQ

### Is GLM-5.3-Flash better than Inkling-Small?

GLM-5.3-Flash is the stronger model overall, scoring 49.6 to 46.5 on the Noometry Index.

### Which is cheaper, GLM-5.3-Flash or Inkling-Small?

GLM-5.3-Flash is cheaper. It lists at $0.15 per million input tokens and $0.50 per million output tokens; Inkling-Small lists at $0.45 and $1.20.

### Is GLM-5.3-Flash or Inkling-Small better for coding?

GLM-5.3-Flash scores higher on coding benchmarks: 53.7 versus 43.3 in the Noometry coding category.

### Which has the bigger context window?

GLM-5.3-Flash does, with 1M tokens against 524K.

### How many benchmarks do GLM-5.3-Flash and Inkling-Small share?

29 benchmarks have published results for both models. GLM-5.3-Flash has 36 scored results on Noometry and Inkling-Small has 33.