# DeepSeek V4 Flash vs GLM-5.1

> DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 47.6 on the Noometry Index.

- Canonical page: https://noometry.com/compare/deepseek-v4-flash-vs-glm-5-1
- Last updated: 2026-10-10
- Shared benchmarks: 33

## Summary

- They share 33 benchmarks with published results for both. DeepSeek V4 Flash scores higher in 3 categories and GLM-5.1 in 5 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek V4 Flash leads 53.7 to 38.2.
- The biggest single-benchmark swing is ProofBench: 56% for DeepSeek V4 Flash and 22.2% for GLM-5.1.
- DeepSeek V4 Flash is cheaper at $0.15 / $0.60 per million input/output tokens, against $1.40 / $4.40 for GLM-5.1.
- DeepSeek V4 Flash accepts more context: 1M tokens versus 200K.

## Snapshot

| | DeepSeek V4 Flash | GLM-5.1 |
|---|---|---|
| Provider | DeepSeek | Z.ai (Zhipu) |
| Noometry Index | 53.6 | 47.6 |
| Rank | 32 | 51 |
| Context | 1M | 200K |
| Input $/M | $0.15 | $1.40 |
| Output $/M | $0.60 | $4.40 |
| Weights | Open | Open |

## Coding

- DeepSeek V4 Flash: 47.6 (#56)
- GLM-5.1: 48.5 (#52)

| Benchmark | DeepSeek V4 Flash | GLM-5.1 |
|---|---|---|
| LMArena WebDev | 1581 | 1509 |
| SciCode | 49.9% | 43.8% |
| WeirdML | 63% | 57.1% |
| LMArena Coding | 1457 | 1486 |
| ALE-Bench | 1,306 | 887.1 |
| SWE-bench Verified | — | 74.2% |
| FrontierCode | 18.8% | — |

## Agentic & Tool Use

- DeepSeek V4 Flash: —
- GLM-5.1: 24.8 (#106)

| Benchmark | DeepSeek V4 Flash | GLM-5.1 |
|---|---|---|
| APEX-Agents | — | 40.9% |
| ExploitBench | — | 18.1% |
| GBAEval | — | 0% |
| Vending-Bench 2 | — | 5,634 |

## Reasoning

- DeepSeek V4 Flash: 53.7 (#25)
- GLM-5.1: 38.2 (#57)

| Benchmark | DeepSeek V4 Flash | GLM-5.1 |
|---|---|---|
| SimpleBench | 61.1% | 55.1% |
| NYT Connections (extended) | 89.6% | 77.7% |
| CritPt | 16.6% | 4.6% |
| Chess Puzzles | 33% | 19% |
| LMArena Hard Prompts | 1443 | 1472 |
| Epoch Capabilities Index | 154.51 | 149.86 |
| ARC-AGI-2 | 61.4% | — |
| Kagi LLM Benchmark | 52.2% | — |
| ARC-AGI-1 | 89% | — |
| Thematic Generalization | — | 69.8% |
| Mystery Game Puzzles | 34% | — |
| DTBench | 90.9% | — |
| LMCA | 41.7% | — |

## Math

- DeepSeek V4 Flash: 60.0 (#31)
- GLM-5.1: 49.0 (#48)

| Benchmark | DeepSeek V4 Flash | GLM-5.1 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 57.5% | 36.8% |
| MathArena Final-Answer Competitions | 76.5% | 67.1% |
| OTIS Mock AIME 2024-2025 | 94.4% | 93.3% |
| ProofBench | 56% | 22.2% |
| LMArena Math | 1427 | 1472 |
| FrontierMath Tier 4 | 24.4% | — |
| FrontierMath (Feb 2025 set) | — | 33.4% |
| FrontierMath Tier 4 (v1) | — | 12.5% |

## Knowledge

- DeepSeek V4 Flash: 55.5 (#43)
- GLM-5.1: 55.0 (#45)

| Benchmark | DeepSeek V4 Flash | GLM-5.1 |
|---|---|---|
| GPQA Diamond | 91% | 89.9% |
| SimpleQA Verified | 33.6% | 34% |
| LMArena Expert | 1441 | 1477 |

## Multilingual

- DeepSeek V4 Flash: 53.2 (#71)
- GLM-5.1: 55.2 (#34)

| Benchmark | DeepSeek V4 Flash | GLM-5.1 |
|---|---|---|
| LMArena Non-English | 1420 | 1448 |
| LMArena Chinese | 1467 | 1516 |
| LMArena French | 1438 | 1478 |
| LMArena German | 1418 | 1466 |
| LMArena Japanese | 1403 | 1436 |
| LMArena Korean | 1383 | 1423 |
| LMArena Russian | 1428 | 1453 |
| LMArena Spanish | 1437 | 1467 |

## Instruction Following

- DeepSeek V4 Flash: 75.1 (#76)
- GLM-5.1: 76.5 (#41)

| Benchmark | DeepSeek V4 Flash | GLM-5.1 |
|---|---|---|
| LMArena Instruction Following | 1421 | 1451 |

## Long Context

- DeepSeek V4 Flash: 43.8 (#81)
- GLM-5.1: 44.9 (#51)

| Benchmark | DeepSeek V4 Flash | GLM-5.1 |
|---|---|---|
| LMArena Longer Query | 1435 | 1467 |

## Writing & Preference

- DeepSeek V4 Flash: 64.0 (#59)
- GLM-5.1: 67.2 (#29)

| Benchmark | DeepSeek V4 Flash | GLM-5.1 |
|---|---|---|
| LMArena Text | 1432 | 1462 |
| LMArena Creative Writing | 1403 | 1454 |
| EQ-Bench Creative Writing | 1559 | 1592 |
| LMArena Multi-Turn | 1449 | 1473 |

## FAQ

### Is DeepSeek V4 Flash better than GLM-5.1?

DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 47.6 on the Noometry Index.

### Which is cheaper, DeepSeek V4 Flash or GLM-5.1?

DeepSeek V4 Flash is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; GLM-5.1 lists at $1.40 and $4.40.

### Is DeepSeek V4 Flash or GLM-5.1 better for coding?

They score almost the same on coding (47.6 vs 48.5); test both on your own repository before choosing.

### Which has the bigger context window?

DeepSeek V4 Flash does, with 1M tokens against 200K.

### How many benchmarks do DeepSeek V4 Flash and GLM-5.1 share?

33 benchmarks have published results for both models. DeepSeek V4 Flash has 41 scored results on Noometry and GLM-5.1 has 41.