Model comparison
GLM-5.3 vs GLM-5V-Turbo
GLM-5.3 is the stronger model overall, scoring 54.6 to 43.1 on the Noometry Index.
Last verified . 17 shared benchmarks.
Summary
- They share 17 benchmarks with published results for both. GLM-5.3 scores higher in 8 categories and GLM-5V-Turbo in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GLM-5.3 leads 62.1 to 36.6.
- GLM-5V-Turbo is cheaper at $1.20 / $4 per million input/output tokens, against $1.40 / $4.40 for GLM-5.3.
- GLM-5.3 accepts more context: 1M tokens versus 200K.
- GLM-5.3 has downloadable open weights; the other is API-only.
Side by side
| GLM-5.3 | GLM-5V-Turbo | |
|---|---|---|
| Provider | Z.ai (Zhipu) | Z.ai (Zhipu) |
| Noometry Index | 54.6 | 43.1 |
| Released | 2026-08-14 | 2026-04-01 |
| Weights | Open | Proprietary |
| Context window | 1M | 200K |
| Max output | 131K | 131K |
| Input $ / M tokens | $1.40 | $1.20 |
| Output $ / M tokens | $4.40 | $4 |
| Results tracked | 41 | 19 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.3 leads
GLM-5.3: 59.0 (#12), GLM-5V-Turbo: 41.8 (#105)
| Benchmark | GLM-5.3 | GLM-5V-Turbo |
|---|---|---|
| LMArena WebDev | 1623 | 1400 |
| LMArena Coding | 1499 | 1467 |
| DeepSWE | 69% | — |
| FrontierCode | 40.1% | — |
| FrontierSWE | 30.2% | — |
| SciCode | 59% | — |
| WeirdML | 75.4% | — |
| ALE-Bench | 1,317 | — |
Agentic & Tool Use Not comparable
GLM-5.3: 36.3 (#34), GLM-5V-Turbo: —
| Benchmark | GLM-5.3 | GLM-5V-Turbo |
|---|---|---|
| APEX-Agents | 56.6% | — |
| Vending-Bench 2 | 8,164 | — |
Reasoning GLM-5.3 leads
GLM-5.3: 45.2 (#42), GLM-5V-Turbo: 28.5 (#87)
| Benchmark | GLM-5.3 | GLM-5V-Turbo |
|---|---|---|
| LMArena Hard Prompts | 1491 | 1443 |
| NYT Connections (extended) | 74.2% | — |
| CritPt | 19.1% | — |
| Chess Puzzles | 21% | — |
| Mystery Game Puzzles | 33% | — |
| DTBench | 87.7% | — |
| LMCA | 55.5% | — |
| Bench to the Future 3 | 0.15 | — |
| Epoch Capabilities Index | 155.56 | — |
Math GLM-5.3 leads
GLM-5.3: 62.1 (#27), GLM-5V-Turbo: 36.6 (#109)
| Benchmark | GLM-5.3 | GLM-5V-Turbo |
|---|---|---|
| LMArena Math | 1493 | 1440 |
| FrontierMath (Tiers 1-3) | 68.8% | — |
| FrontierMath Tier 4 | 29.3% | — |
| OTIS Mock AIME 2024-2025 | 91.1% | — |
| ProofBench | 49% | — |
Knowledge GLM-5.3 leads
GLM-5.3: 58.3 (#34), GLM-5V-Turbo: 39.9 (#113)
| Benchmark | GLM-5.3 | GLM-5V-Turbo |
|---|---|---|
| LMArena Expert | 1512 | 1451 |
| GPQA Diamond | 90.9% | — |
| SimpleQA Verified | 41% | — |
Multimodal Not comparable
GLM-5.3: —, GLM-5V-Turbo: 40.1 (#41)
| Benchmark | GLM-5.3 | GLM-5V-Turbo |
|---|---|---|
| LMArena Vision | — | 1264 |
| LMArena Document | — | 1416 |
Multilingual GLM-5.3 leads
GLM-5.3: 55.9 (#27), GLM-5V-Turbo: 53.2 (#69)
| Benchmark | GLM-5.3 | GLM-5V-Turbo |
|---|---|---|
| LMArena Non-English | 1457 | 1420 |
| LMArena Chinese | 1523 | 1488 |
| LMArena French | 1499 | 1443 |
| LMArena German | 1505 | 1422 |
| LMArena Korean | 1478 | 1399 |
| LMArena Russian | 1460 | 1432 |
| LMArena Spanish | 1444 | 1446 |
| LMArena Japanese | 1438 | — |
Instruction Following GLM-5.3 leads
GLM-5.3: 77.8 (#15), GLM-5V-Turbo: 75.1 (#75)
| Benchmark | GLM-5.3 | GLM-5V-Turbo |
|---|---|---|
| LMArena Instruction Following | 1481 | 1422 |
Long Context GLM-5.3 leads
GLM-5.3: 45.5 (#32), GLM-5V-Turbo: 44.0 (#76)
| Benchmark | GLM-5.3 | GLM-5V-Turbo |
|---|---|---|
| LMArena Longer Query | 1485 | 1439 |
Writing & Preference GLM-5.3 leads
GLM-5.3: 75.9 (#6), GLM-5V-Turbo: 62.6 (#69)
| Benchmark | GLM-5.3 | GLM-5V-Turbo |
|---|---|---|
| LMArena Text | 1473 | 1437 |
| LMArena Creative Writing | 1459 | 1414 |
| LMArena Multi-Turn | 1473 | 1432 |
| EQ-Bench Creative Writing | 2075 | — |
Frequently asked questions
Is GLM-5.3 better than GLM-5V-Turbo?
GLM-5.3 is the stronger model overall, scoring 54.6 to 43.1 on the Noometry Index.
Which is cheaper, GLM-5.3 or GLM-5V-Turbo?
GLM-5V-Turbo is cheaper. It lists at $1.20 per million input tokens and $4 per million output tokens; GLM-5.3 lists at $1.40 and $4.40.
Is GLM-5.3 or GLM-5V-Turbo better for coding?
GLM-5.3 scores higher on coding benchmarks: 59.0 versus 41.8 in the Noometry coding category.
Which has the bigger context window?
GLM-5.3 does, with 1M tokens against 200K.
How many benchmarks do GLM-5.3 and GLM-5V-Turbo share?
17 benchmarks have published results for both models. GLM-5.3 has 41 scored results on Noometry and GLM-5V-Turbo has 19.