Model comparison
GLM-4.6 vs GLM-5.3-Flash
GLM-5.3-Flash is the stronger model overall, scoring 49.6 to 40.9 on the Noometry Index.
Last verified . 21 shared benchmarks.
Summary
- They share 21 benchmarks with published results for both. GLM-4.6 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 knowledge, where GLM-5.3-Flash leads 58.4 to 39.6.
- The biggest single-benchmark swing is CritPt: 1.1% for GLM-4.6 and 15.4% 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-4.6.
- GLM-5.3-Flash accepts more context: 1M tokens versus 205K.
Side by side
| GLM-4.6 | GLM-5.3-Flash | |
|---|---|---|
| Provider | Z.ai (Zhipu) | Z.ai (Zhipu) |
| Noometry Index | 40.9 | 49.6 |
| Released | 2025-09-30 | 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 | 29 | 36 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.3-Flash leads
GLM-4.6: 39.9 (#137), GLM-5.3-Flash: 53.7 (#26)
| Benchmark | GLM-4.6 | GLM-5.3-Flash |
|---|---|---|
| LMArena WebDev | 1339 | 1614 |
| SciCode | 38.4% | 51.6% |
| LMArena Coding | 1449 | 1506 |
| ALE-Bench | 340.82 | 303.55 |
| DeepSWE | — | 63.4% |
| FrontierCode | — | 31.8% |
| SWE-bench Verified (bash only) | 55.4% | — |
Agentic & Tool Use GLM-5.3-Flash leads
GLM-4.6: 32.1 (#60), GLM-5.3-Flash: 34.3 (#41)
| Benchmark | GLM-4.6 | GLM-5.3-Flash |
|---|---|---|
| Terminal-Bench | 24.5% | — |
| APEX-Agents | — | 52.8% |
| Berkeley Function Calling Leaderboard | 72.4% | — |
| GDP.pdf | — | 14% |
Reasoning GLM-5.3-Flash leads
GLM-4.6: 23.1 (#168), GLM-5.3-Flash: 31.7 (#69)
| Benchmark | GLM-4.6 | GLM-5.3-Flash |
|---|---|---|
| CritPt | 1.1% | 15.4% |
| LMArena Hard Prompts | 1440 | 1490 |
| Kagi LLM Benchmark | 47.4% | — |
| Chess Puzzles | — | 14% |
| Mystery Game Puzzles | — | 8% |
| Surface Evolver Bench | — | 52.5% |
| Bench to the Future 3 | — | 0.15 |
| Epoch Capabilities Index | — | 151.9 |
Math GLM-5.3-Flash leads
GLM-4.6: 36.4 (#115), GLM-5.3-Flash: 53.1 (#38)
| Benchmark | GLM-4.6 | GLM-5.3-Flash |
|---|---|---|
| LMArena Math | 1431 | 1499 |
| FrontierMath (Tiers 1-3) | — | 55.8% |
| FrontierMath Tier 4 | — | 17.1% |
| OTIS Mock AIME 2024-2025 | — | 93.9% |
| ProofBench | — | 21% |
| FrontierMath (Feb 2025 set) | 3.8% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge GLM-5.3-Flash leads
GLM-4.6: 39.6 (#118), GLM-5.3-Flash: 58.4 (#33)
| Benchmark | GLM-4.6 | GLM-5.3-Flash |
|---|---|---|
| LMArena Expert | 1432 | 1512 |
| GPQA Diamond | — | 90.2% |
| Vectara Hallucination Rate | 9.5% | — |
Multimodal Not comparable
GLM-4.6: —, GLM-5.3-Flash: 42.6 (#25)
| Benchmark | GLM-4.6 | GLM-5.3-Flash |
|---|---|---|
| LMArena Vision | — | 1304 |
Multilingual GLM-5.3-Flash leads
GLM-4.6: 53.6 (#62), GLM-5.3-Flash: 56.2 (#24)
| Benchmark | GLM-4.6 | GLM-5.3-Flash |
|---|---|---|
| LMArena Non-English | 1426 | 1462 |
| LMArena Chinese | 1500 | 1529 |
| LMArena French | 1459 | 1495 |
| LMArena German | 1447 | 1457 |
| LMArena Japanese | 1394 | 1420 |
| LMArena Korean | 1401 | 1443 |
| LMArena Russian | 1418 | 1472 |
| LMArena Spanish | 1436 | 1471 |
Instruction Following GLM-5.3-Flash leads
GLM-4.6: 74.5 (#93), GLM-5.3-Flash: 77.8 (#17)
| Benchmark | GLM-4.6 | GLM-5.3-Flash |
|---|---|---|
| LMArena Instruction Following | 1410 | 1480 |
Long Context GLM-5.3-Flash leads
GLM-4.6: 43.4 (#91), GLM-5.3-Flash: 45.3 (#41)
| Benchmark | GLM-4.6 | GLM-5.3-Flash |
|---|---|---|
| LMArena Longer Query | 1423 | 1479 |
Writing & Preference GLM-5.3-Flash leads
GLM-4.6: 61.3 (#86), GLM-5.3-Flash: 65.4 (#49)
| Benchmark | GLM-4.6 | GLM-5.3-Flash |
|---|---|---|
| LMArena Text | 1440 | 1471 |
| LMArena Creative Writing | 1411 | 1440 |
| LMArena Multi-Turn | 1427 | 1471 |
| EQ-Bench Creative Writing | 1411 | — |
Frequently asked questions
Is GLM-4.6 better than GLM-5.3-Flash?
GLM-5.3-Flash is the stronger model overall, scoring 49.6 to 40.9 on the Noometry Index.
Which is cheaper, GLM-4.6 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-4.6 lists at $0.60 and $2.20.
Is GLM-4.6 or GLM-5.3-Flash better for coding?
GLM-5.3-Flash scores higher on coding benchmarks: 53.7 versus 39.9 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-4.6 and GLM-5.3-Flash share?
21 benchmarks have published results for both models. GLM-4.6 has 29 scored results on Noometry and GLM-5.3-Flash has 36.