Model comparison
DeepSeek V4 Flash vs GLM-5.3-Flash
DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 51.8 on the Noometry Index.
Last verified . 32 shared benchmarks.
Summary
- They share 32 benchmarks with published results for both. DeepSeek V4 Flash scores higher in 2 categories and GLM-5.3-Flash in 6 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where DeepSeek V4 Flash leads 60.3 to 53.3.
- The biggest single-benchmark swing is ProofBench: 56% for DeepSeek V4 Flash and 21% for GLM-5.3-Flash.
- GLM-5.3-Flash is cheaper at $0.15 / $0.50 per million input/output tokens, against $0.15 / $0.60 for DeepSeek V4 Flash.
Side by side
| DeepSeek V4 Flash | GLM-5.3-Flash | |
|---|---|---|
| Provider | DeepSeek | Z.ai (Zhipu) |
| Noometry Index | 53.6 | 51.8 |
| Released | 2026-04-24 | 2026-08-20 |
| Weights | Open | Open |
| Context window | 1M | 1M |
| Max output | 393K | 131K |
| Input $ / M tokens | $0.15 | $0.15 |
| Output $ / M tokens | $0.60 | $0.50 |
| Results tracked | 41 | 40 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.3-Flash leads
DeepSeek V4 Flash: 47.9 (#59), GLM-5.3-Flash: 53.1 (#31)
| Benchmark | DeepSeek V4 Flash | GLM-5.3-Flash |
|---|---|---|
| FrontierCode | 18.8% | 31.8% |
| LMArena WebDev | 1582 | 1609 |
| SciCode | 49.9% | 51.6% |
| LMArena Coding | 1457 | 1508 |
| ALE-Bench | 1,306 | 303.55 |
| DeepSWE | — | 63.4% |
| CursorBench | — | 36.8% |
| FrontierSWE | — | 18.1% |
| WeirdML | 63% | — |
Agentic & Tool Use Not comparable
DeepSeek V4 Flash: —, GLM-5.3-Flash: 34.2 (#47)
| Benchmark | DeepSeek V4 Flash | GLM-5.3-Flash |
|---|---|---|
| APEX-Agents | — | 52.8% |
| GDP.pdf | — | 14% |
Reasoning DeepSeek V4 Flash leads
DeepSeek V4 Flash: 53.7 (#30), GLM-5.3-Flash: 48.0 (#42)
| Benchmark | DeepSeek V4 Flash | GLM-5.3-Flash |
|---|---|---|
| ARC-AGI-2 | 61.4% | 65.8% |
| ARC-AGI-1 | 89% | 91% |
| CritPt | 16.6% | 15.4% |
| Chess Puzzles | 33% | 14% |
| LMArena Hard Prompts | 1444 | 1491 |
| Mystery Game Puzzles | 34% | 8% |
| Epoch Capabilities Index | 154.49 | 151.88 |
| SimpleBench | 61.1% | — |
| Kagi LLM Benchmark | 52.2% | — |
| NYT Connections (extended) | 89.6% | — |
| DTBench | 90.9% | — |
| LMCA | 41.7% | — |
| Surface Evolver Bench | — | 52.5% |
| Bench to the Future 3 | — | 0.15 |
Math DeepSeek V4 Flash leads
DeepSeek V4 Flash: 60.3 (#37), GLM-5.3-Flash: 53.3 (#47)
| Benchmark | DeepSeek V4 Flash | GLM-5.3-Flash |
|---|---|---|
| FrontierMath (Tiers 1-3) | 57.5% | 55.8% |
| FrontierMath Tier 4 | 24.4% | 17.1% |
| OTIS Mock AIME 2024-2025 | 94.4% | 93.9% |
| ProofBench | 56% | 21% |
| LMArena Math | 1427 | 1500 |
| MathArena Final-Answer Competitions | 76.5% | — |
Knowledge GLM-5.3-Flash leads
DeepSeek V4 Flash: 55.4 (#48), GLM-5.3-Flash: 58.4 (#36)
| Benchmark | DeepSeek V4 Flash | GLM-5.3-Flash |
|---|---|---|
| GPQA Diamond | 91% | 90.2% |
| LMArena Expert | 1441 | 1513 |
| SimpleQA Verified | 33.6% | — |
Multimodal Not comparable
DeepSeek V4 Flash: —, GLM-5.3-Flash: 42.8 (#27)
| Benchmark | DeepSeek V4 Flash | GLM-5.3-Flash |
|---|---|---|
| LMArena Vision | — | 1296 |
Multilingual GLM-5.3-Flash leads
DeepSeek V4 Flash: 53.0 (#72), GLM-5.3-Flash: 56.0 (#25)
| Benchmark | DeepSeek V4 Flash | GLM-5.3-Flash |
|---|---|---|
| LMArena Non-English | 1420 | 1462 |
| LMArena Chinese | 1468 | 1527 |
| LMArena French | 1439 | 1496 |
| LMArena German | 1418 | 1470 |
| LMArena Japanese | 1406 | 1429 |
| LMArena Korean | 1384 | 1446 |
| LMArena Russian | 1428 | 1469 |
| LMArena Spanish | 1436 | 1471 |
Instruction Following GLM-5.3-Flash leads
DeepSeek V4 Flash: 74.9 (#81), GLM-5.3-Flash: 77.5 (#20)
| Benchmark | DeepSeek V4 Flash | GLM-5.3-Flash |
|---|---|---|
| LMArena Instruction Following | 1421 | 1478 |
Long Context GLM-5.3-Flash leads
DeepSeek V4 Flash: 43.8 (#85), GLM-5.3-Flash: 45.4 (#39)
| Benchmark | DeepSeek V4 Flash | GLM-5.3-Flash |
|---|---|---|
| LMArena Longer Query | 1434 | 1482 |
Writing & Preference GLM-5.3-Flash leads
DeepSeek V4 Flash: 63.8 (#61), GLM-5.3-Flash: 65.3 (#50)
| Benchmark | DeepSeek V4 Flash | GLM-5.3-Flash |
|---|---|---|
| LMArena Text | 1432 | 1471 |
| LMArena Creative Writing | 1403 | 1442 |
| LMArena Multi-Turn | 1449 | 1467 |
| EQ-Bench Creative Writing | 1559 | — |
Frequently asked questions
Is DeepSeek V4 Flash better than GLM-5.3-Flash?
DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 51.8 on the Noometry Index.
Which is cheaper, DeepSeek V4 Flash or GLM-5.3-Flash?
GLM-5.3-Flash is cheaper. It lists at $0.15 per million input tokens and $0.50 per million output tokens; DeepSeek V4 Flash lists at $0.15 and $0.60.
Is DeepSeek V4 Flash or GLM-5.3-Flash better for coding?
GLM-5.3-Flash scores higher on coding benchmarks: 53.1 versus 47.9 in the Noometry coding category.
Which has the bigger context window?
Both accept 1M tokens.
How many benchmarks do DeepSeek V4 Flash and GLM-5.3-Flash share?
32 benchmarks have published results for both models. DeepSeek V4 Flash has 41 scored results on Noometry and GLM-5.3-Flash has 40.