# Gemini 3.1 Pro Preview vs GLM-5.2

> Gemini 3.1 Pro Preview is the stronger model overall, scoring 56.7 to 51.1 on the Noometry Index. GLM-5.2 costs 2.1× less per token, which makes it the better buy when Gemini 3.1 Pro Preview's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gemini-3-1-pro-preview-vs-glm-5-2
- Last updated: 2026-10-10
- Shared benchmarks: 48

## Summary

- They share 48 benchmarks with published results for both. Gemini 3.1 Pro Preview scores higher in 7 categories and GLM-5.2 in 2 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3.1 Pro Preview leads 71.7 to 42.3.
- The biggest single-benchmark swing is ARC-AGI-2: 77.1% for Gemini 3.1 Pro Preview and 22.8% for GLM-5.2.
- GLM-5.2 is cheaper at $1.40 / $4.40 per million input/output tokens, against $2 / $12 for Gemini 3.1 Pro Preview.
- Gemini 3.1 Pro Preview accepts more context: 1.05M tokens versus 1M.
- GLM-5.2 has downloadable open weights; the other is API-only.

## Snapshot

| | Gemini 3.1 Pro Preview | GLM-5.2 |
|---|---|---|
| Provider | Google | Z.ai (Zhipu) |
| Noometry Index | 56.7 | 51.1 |
| Rank | 23 | 44 |
| Context | 1.05M | 1M |
| Input $/M | $2 | $1.40 |
| Output $/M | $12 | $4.40 |
| Weights | Proprietary | Open |

## Coding

- Gemini 3.1 Pro Preview: 42.5 (#99)
- GLM-5.2: 51.3 (#41)

| Benchmark | Gemini 3.1 Pro Preview | GLM-5.2 |
|---|---|---|
| SWE-bench Verified | 75.6% | 78.7% |
| DeepSWE | 11.7% | 43.8% |
| LMArena WebDev | 1447 | 1603 |
| SciCode | 58.9% | 50.5% |
| WeirdML | 72.1% | 70.1% |
| LMArena Coding | 1484 | 1485 |
| ALE-Bench | 1,161 | 1,047 |
| FrontierCode | — | 24.5% |
| GSO | 22.6% | — |
| MirrorCode | 8.9% | — |
| AlgoTune | 2.02 | — |

## Agentic & Tool Use

- Gemini 3.1 Pro Preview: 37.7 (#34)
- GLM-5.2: 32.4 (#63)

| Benchmark | Gemini 3.1 Pro Preview | GLM-5.2 |
|---|---|---|
| APEX-Agents | 35.3% | 45.2% |
| τ²-bench Banking | 26% | 37.1% |
| PostTrainBench | 22% | 31.7% |
| GBAEval | 0.8% | 0% |
| Vending-Bench 2 | 3,774 | 8,314 |
| Terminal-Bench | 80.2% | — |
| DeepResearch Bench | 47.8% | — |
| BALROG | 57% | — |
| ExploitBench | 26.1% | — |
| GDP.pdf | 17% | — |
| LMArena Search | 1211 | — |
| METR Time Horizons | 77% | — |

## Reasoning

- Gemini 3.1 Pro Preview: 71.7 (#12)
- GLM-5.2: 42.3 (#52)

| Benchmark | Gemini 3.1 Pro Preview | GLM-5.2 |
|---|---|---|
| ARC-AGI-2 | 77.1% | 22.8% |
| SimpleBench | 79.6% | 58.8% |
| NYT Connections (extended) | 97.4% | 74.3% |
| ARC-AGI-1 | 98% | 77% |
| CritPt | 17.7% | 20.9% |
| Chess Puzzles | 55% | 21% |
| EBR-Bench | 14.3% | 9.5% |
| LMArena Hard Prompts | 1485 | 1480 |
| Mystery Game Puzzles | 34% | 19% |
| DTBench | 97.1% | 93.6% |
| LMCA | 53.8% | 45.8% |
| Epoch Capabilities Index | 154.77 | 151.78 |
| Kagi LLM Benchmark | — | 62.6% |
| EnigmaEval | 36.8% | — |
| Thematic Generalization | 79.4% | — |
| Surface Evolver Bench | — | 55.6% |
| ForecastBench | 59 | — |

## Math

- Gemini 3.1 Pro Preview: 62.1 (#34)
- GLM-5.2: 55.7 (#43)

| Benchmark | Gemini 3.1 Pro Preview | GLM-5.2 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 59.6% | 59.2% |
| FrontierMath Tier 4 | 26.8% | 29.3% |
| MathArena Final-Answer Competitions | 86.5% | 67.6% |
| OTIS Mock AIME 2024-2025 | 95.6% | 86.4% |
| ProofBench | 26% | 35% |
| LMArena Math | 1485 | 1482 |
| FrontierMath (Feb 2025 set) | 36.9% | — |
| FrontierMath Tier 4 (v1) | 16.7% | — |

## Knowledge

- Gemini 3.1 Pro Preview: 71.8 (#3)
- GLM-5.2: 57.1 (#40)

| Benchmark | Gemini 3.1 Pro Preview | GLM-5.2 |
|---|---|---|
| GPQA Diamond | 94.4% | 91.9% |
| SimpleQA Verified | 73.5% | 34.2% |
| LMArena Expert | 1485 | 1486 |
| Humanity's Last Exam | 46.4% | — |
| Vectara Hallucination Rate | 10.4% | — |

## Multimodal

- Gemini 3.1 Pro Preview: 37.9 (#69)
- GLM-5.2: —

| Benchmark | Gemini 3.1 Pro Preview | GLM-5.2 |
|---|---|---|
| LMArena Vision | 1296 | — |
| Blueprint-Bench 2 | 26.5% | — |
| Furniture Assembly | 26.7% | — |
| LMArena Document | 1444 | — |

## Multilingual

- Gemini 3.1 Pro Preview: 57.0 (#12)
- GLM-5.2: 55.8 (#26)

| Benchmark | Gemini 3.1 Pro Preview | GLM-5.2 |
|---|---|---|
| LMArena Non-English | 1477 | 1459 |
| LMArena Chinese | 1529 | 1519 |
| LMArena French | 1487 | 1479 |
| LMArena German | 1491 | 1468 |
| LMArena Japanese | 1493 | 1451 |
| LMArena Korean | 1455 | 1445 |
| LMArena Russian | 1498 | 1466 |
| LMArena Spanish | 1479 | 1477 |

## Instruction Following

- Gemini 3.1 Pro Preview: 77.0 (#32)
- GLM-5.2: 76.9 (#34)

| Benchmark | Gemini 3.1 Pro Preview | GLM-5.2 |
|---|---|---|
| LMArena Instruction Following | 1466 | 1465 |

## Long Context

- Gemini 3.1 Pro Preview: 47.4 (#18)
- GLM-5.2: 45.3 (#43)

| Benchmark | Gemini 3.1 Pro Preview | GLM-5.2 |
|---|---|---|
| LMArena Longer Query | 1483 | 1479 |
| CL-bench | 20.8% | — |
| CL-bench Life | 16.9% | — |

## Writing & Preference

- Gemini 3.1 Pro Preview: 66.1 (#37)
- GLM-5.2: 70.4 (#21)

| Benchmark | Gemini 3.1 Pro Preview | GLM-5.2 |
|---|---|---|
| LMArena Text | 1481 | 1470 |
| LMArena Creative Writing | 1482 | 1462 |
| EQ-Bench Creative Writing | 1491 | 1757 |
| EQ-Bench 4 | 1142 | 1222 |
| LMArena Multi-Turn | 1488 | 1469 |

## FAQ

### Is Gemini 3.1 Pro Preview better than GLM-5.2?

Gemini 3.1 Pro Preview is the stronger model overall, scoring 56.7 to 51.1 on the Noometry Index. GLM-5.2 costs 2.1× less per token, which makes it the better buy when Gemini 3.1 Pro Preview's lead doesn't matter for your workload.

### Which is cheaper, Gemini 3.1 Pro Preview or GLM-5.2?

GLM-5.2 is cheaper. It lists at $1.40 per million input tokens and $4.40 per million output tokens; Gemini 3.1 Pro Preview lists at $2 and $12.

### Is Gemini 3.1 Pro Preview or GLM-5.2 better for coding?

GLM-5.2 scores higher on coding benchmarks: 51.3 versus 42.5 in the Noometry coding category.

### Which has the bigger context window?

Gemini 3.1 Pro Preview does, with 1.05M tokens against 1M.

### How many benchmarks do Gemini 3.1 Pro Preview and GLM-5.2 share?

48 benchmarks have published results for both models. Gemini 3.1 Pro Preview has 71 scored results on Noometry and GLM-5.2 has 51.