Model comparison
GLM-4.7 vs GLM-5.3-Flash
GLM-5.3-Flash is the stronger model overall, scoring 51.8 to 42.0 on the Noometry Index.
Last verified . 26 shared benchmarks.
Summary
- They share 26 benchmarks with published results for both. GLM-4.7 scores higher in 0 categories and GLM-5.3-Flash in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GLM-5.3-Flash leads 48.0 to 24.3.
- The biggest single-benchmark swing is ProofBench: 6% for GLM-4.7 and 21% for GLM-5.3-Flash.
- GLM-5.3-Flash is cheaper at $0.15 / $0.50 per million input/output tokens, against $0.60 / $2.20 for GLM-4.7.
- GLM-5.3-Flash accepts more context: 1M tokens versus 205K.
Side by side
| GLM-4.7 | GLM-5.3-Flash | |
|---|---|---|
| Provider | Z.ai (Zhipu) | Z.ai (Zhipu) |
| Noometry Index | 42.0 | 51.8 |
| Released | 2025-12-22 | 2026-08-20 |
| Weights | Open | Open |
| Context window | 205K | 1M |
| Max output | 131K | 131K |
| Input $ / M tokens | $0.60 | $0.15 |
| Output $ / M tokens | $2.20 | $0.50 |
| Results tracked | 36 | 40 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.3-Flash leads
GLM-4.7: 44.0 (#79), GLM-5.3-Flash: 53.1 (#31)
| Benchmark | GLM-4.7 | GLM-5.3-Flash |
|---|---|---|
| LMArena WebDev | 1435 | 1609 |
| SciCode | 45.1% | 51.6% |
| LMArena Coding | 1454 | 1508 |
| ALE-Bench | 399.48 | 303.55 |
| DeepSWE | — | 63.4% |
| FrontierCode | — | 31.8% |
| CursorBench | — | 36.8% |
| FrontierSWE | — | 18.1% |
Agentic & Tool Use GLM-5.3-Flash leads
GLM-4.7: 26.5 (#103), GLM-5.3-Flash: 34.2 (#47)
| Benchmark | GLM-4.7 | GLM-5.3-Flash |
|---|---|---|
| Terminal-Bench | 33.4% | — |
| APEX-Agents | — | 52.8% |
| GDP.pdf | — | 14% |
| Vending-Bench 2 | 2,377 | — |
Reasoning GLM-5.3-Flash leads
GLM-4.7: 24.3 (#164), GLM-5.3-Flash: 48.0 (#42)
| Benchmark | GLM-4.7 | GLM-5.3-Flash |
|---|---|---|
| CritPt | 1.7% | 15.4% |
| Chess Puzzles | 6% | 14% |
| LMArena Hard Prompts | 1443 | 1491 |
| Epoch Capabilities Index | 143.51 | 151.88 |
| ARC-AGI-2 | — | 65.8% |
| SimpleBench | 47.7% | — |
| ARC-AGI-1 | — | 91% |
| Mystery Game Puzzles | — | 8% |
| Surface Evolver Bench | — | 52.5% |
| Bench to the Future 3 | — | 0.15 |
Math GLM-5.3-Flash leads
GLM-4.7: 38.6 (#135), GLM-5.3-Flash: 53.3 (#47)
| Benchmark | GLM-4.7 | GLM-5.3-Flash |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 83.3% | 93.9% |
| ProofBench | 6% | 21% |
| LMArena Math | 1423 | 1500 |
| FrontierMath (Tiers 1-3) | — | 55.8% |
| FrontierMath Tier 4 | — | 17.1% |
| FrontierMath (Feb 2025 set) | 2.4% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge GLM-5.3-Flash leads
GLM-4.7: 47.0 (#80), GLM-5.3-Flash: 58.4 (#36)
| Benchmark | GLM-4.7 | GLM-5.3-Flash |
|---|---|---|
| GPQA Diamond | 83.3% | 90.2% |
| LMArena Expert | 1424 | 1513 |
| SimpleQA Verified | 32.2% | — |
| Vectara Hallucination Rate | 11.7% | — |
Multimodal Not comparable
GLM-4.7: —, GLM-5.3-Flash: 42.8 (#27)
| Benchmark | GLM-4.7 | GLM-5.3-Flash |
|---|---|---|
| LMArena Vision | — | 1296 |
Multilingual GLM-5.3-Flash leads
GLM-4.7: 52.8 (#79), GLM-5.3-Flash: 56.0 (#25)
| Benchmark | GLM-4.7 | GLM-5.3-Flash |
|---|---|---|
| LMArena Non-English | 1417 | 1462 |
| LMArena Chinese | 1495 | 1527 |
| LMArena French | 1432 | 1496 |
| LMArena German | 1424 | 1470 |
| LMArena Japanese | 1439 | 1429 |
| LMArena Korean | 1399 | 1446 |
| LMArena Russian | 1423 | 1469 |
| LMArena Spanish | 1434 | 1471 |
Instruction Following GLM-5.3-Flash leads
GLM-4.7: 74.4 (#95), GLM-5.3-Flash: 77.5 (#20)
| Benchmark | GLM-4.7 | GLM-5.3-Flash |
|---|---|---|
| LMArena Instruction Following | 1411 | 1478 |
Long Context GLM-5.3-Flash leads
GLM-4.7: 42.8 (#116), GLM-5.3-Flash: 45.4 (#39)
| Benchmark | GLM-4.7 | GLM-5.3-Flash |
|---|---|---|
| LMArena Longer Query | 1432 | 1482 |
| CL-bench | 15.9% | — |
| CL-bench Life | 10.9% | — |
Writing & Preference GLM-5.3-Flash leads
GLM-4.7: 60.9 (#93), GLM-5.3-Flash: 65.3 (#50)
| Benchmark | GLM-4.7 | GLM-5.3-Flash |
|---|---|---|
| LMArena Text | 1435 | 1471 |
| LMArena Creative Writing | 1401 | 1442 |
| LMArena Multi-Turn | 1446 | 1467 |
| EQ-Bench Creative Writing | 1413 | — |
Frequently asked questions
Is GLM-4.7 better than GLM-5.3-Flash?
GLM-5.3-Flash is the stronger model overall, scoring 51.8 to 42.0 on the Noometry Index.
Which is cheaper, GLM-4.7 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; GLM-4.7 lists at $0.60 and $2.20.
Is GLM-4.7 or GLM-5.3-Flash better for coding?
GLM-5.3-Flash scores higher on coding benchmarks: 53.1 versus 44.0 in the Noometry coding category.
Which has the bigger context window?
GLM-5.3-Flash does, with 1M tokens against 205K.
How many benchmarks do GLM-4.7 and GLM-5.3-Flash share?
26 benchmarks have published results for both models. GLM-4.7 has 36 scored results on Noometry and GLM-5.3-Flash has 40.