# GLM-4.7 vs GLM-5

> GLM-5 is the stronger model overall, scoring 45.9 to 41.7 on the Noometry Index. GLM-4.7 costs 1.6× less per token, which makes it the better buy when GLM-5's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/glm-4-7-vs-glm-5
- Last updated: 2026-10-10
- Shared benchmarks: 31

## Summary

- They share 31 benchmarks with published results for both. GLM-4.7 scores higher in 0 categories and GLM-5 in 9 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where GLM-5 leads 44.8 to 37.7.
- The biggest single-benchmark swing is Terminal-Bench: 33.4% for GLM-4.7 and 52.4% for GLM-5.
- GLM-4.7 is cheaper at $0.60 / $2.20 per million input/output tokens, against $1 / $3.20 for GLM-5.

## Snapshot

| | GLM-4.7 | GLM-5 |
|---|---|---|
| Provider | Z.ai (Zhipu) | Z.ai (Zhipu) |
| Noometry Index | 41.7 | 45.9 |
| Rank | 117 | 59 |
| Context | 205K | 205K |
| Input $/M | $0.60 | $1 |
| Output $/M | $2.20 | $3.20 |
| Weights | Open | Open |

## Coding

- GLM-4.7: 43.6 (#74)
- GLM-5: 49.0 (#47)

| Benchmark | GLM-4.7 | GLM-5 |
|---|---|---|
| LMArena WebDev | 1435 | 1434 |
| LMArena Coding | 1454 | 1461 |
| ALE-Bench | 399.48 | 765.62 |
| SWE-bench Verified | — | 72.1% |
| SWE-bench Verified (bash only) | — | 72.8% |
| SWE-bench Multilingual | — | 69.7% |
| SciCode | 45.1% | — |
| WeirdML | — | 48.2% |

## Agentic & Tool Use

- GLM-4.7: 26.4 (#96)
- GLM-5: 30.9 (#65)

| Benchmark | GLM-4.7 | GLM-5 |
|---|---|---|
| Terminal-Bench | 33.4% | 52.4% |
| Vending-Bench 2 | 2,377 | 4,432 |
| τ²-bench Airline | — | 82.5% |
| τ²-bench Banking | — | 9.8% |
| τ²-bench Retail | — | 73.7% |
| τ²-bench Telecom | — | 86.8% |

## Reasoning

- GLM-4.7: 23.7 (#161)
- GLM-5: 27.6 (#100)

| Benchmark | GLM-4.7 | GLM-5 |
|---|---|---|
| SimpleBench | 47.7% | 53.2% |
| Chess Puzzles | 6% | 10% |
| LMArena Hard Prompts | 1443 | 1451 |
| Epoch Capabilities Index | 143.53 | 145.84 |
| ARC-AGI-2 | — | 4.9% |
| Kagi LLM Benchmark | — | 75% |
| NYT Connections (extended) | — | 74.8% |
| ARC-AGI-1 | — | 44.7% |
| CritPt | 1.7% | — |
| ForecastBench | — | 61 |

## Math

- GLM-4.7: 37.7 (#92)
- GLM-5: 44.8 (#66)

| Benchmark | GLM-4.7 | GLM-5 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 83.3% | 80% |
| LMArena Math | 1423 | 1440 |
| FrontierMath (Feb 2025 set) | 2.4% | 16.4% |
| FrontierMath Tier 4 (v1) | 0% | 2.1% |
| MathArena Final-Answer Competitions | — | 65.7% |
| ProofBench | 6% | — |

## Knowledge

- GLM-4.7: 47.0 (#74)
- GLM-5: 52.2 (#59)

| Benchmark | GLM-4.7 | GLM-5 |
|---|---|---|
| GPQA Diamond | 83.3% | 87.8% |
| Vectara Hallucination Rate | 11.7% | 10.1% |
| LMArena Expert | 1424 | 1454 |
| SimpleQA Verified | 32.2% | — |

## Multilingual

- GLM-4.7: 53.0 (#75)
- GLM-5: 53.9 (#57)

| Benchmark | GLM-4.7 | GLM-5 |
|---|---|---|
| LMArena Non-English | 1417 | 1430 |
| LMArena Chinese | 1495 | 1511 |
| LMArena French | 1432 | 1455 |
| LMArena German | 1426 | 1445 |
| LMArena Japanese | 1439 | 1416 |
| LMArena Korean | 1401 | 1421 |
| LMArena Russian | 1423 | 1436 |
| LMArena Spanish | 1433 | 1452 |

## Instruction Following

- GLM-4.7: 74.6 (#90)
- GLM-5: 75.4 (#64)

| Benchmark | GLM-4.7 | GLM-5 |
|---|---|---|
| LMArena Instruction Following | 1411 | 1427 |

## Long Context

- GLM-4.7: 42.8 (#111)
- GLM-5: 44.6 (#58)

| Benchmark | GLM-4.7 | GLM-5 |
|---|---|---|
| CL-bench | 15.9% | 18.7% |
| LMArena Longer Query | 1432 | 1446 |
| CL-bench Life | 10.9% | — |

## Writing & Preference

- GLM-4.7: 61.1 (#89)
- GLM-5: 66.2 (#38)

| Benchmark | GLM-4.7 | GLM-5 |
|---|---|---|
| LMArena Text | 1436 | 1446 |
| LMArena Creative Writing | 1401 | 1438 |
| EQ-Bench Creative Writing | 1413 | 1601 |
| LMArena Multi-Turn | 1446 | 1456 |

## FAQ

### Is GLM-4.7 better than GLM-5?

GLM-5 is the stronger model overall, scoring 45.9 to 41.7 on the Noometry Index. GLM-4.7 costs 1.6× less per token, which makes it the better buy when GLM-5's lead doesn't matter for your workload.

### Which is cheaper, GLM-4.7 or GLM-5?

GLM-4.7 is cheaper. It lists at $0.60 per million input tokens and $2.20 per million output tokens; GLM-5 lists at $1 and $3.20.

### Is GLM-4.7 or GLM-5 better for coding?

GLM-5 scores higher on coding benchmarks: 49.0 versus 43.6 in the Noometry coding category.

### Which has the bigger context window?

Both accept 205K tokens.

### How many benchmarks do GLM-4.7 and GLM-5 share?

31 benchmarks have published results for both models. GLM-4.7 has 36 scored results on Noometry and GLM-5 has 45.