Model comparison
GLM-4.5 vs GPT-5.6 Luna
GPT-5.6 Luna is the stronger model overall, scoring 54.7 to 41.4 on the Noometry Index.
Last verified . 21 shared benchmarks.
Summary
- They share 21 benchmarks with published results for both. GLM-4.5 scores higher in 0 categories and GPT-5.6 Luna in 8 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Luna leads 77.5 to 36.3.
- The biggest single-benchmark swing is WeirdML: 40.6% for GLM-4.5 and 60.9% for GPT-5.6 Luna.
- GPT-5.6 Luna is cheaper at $0.20 / $1.20 per million input/output tokens, against $0.60 / $2.20 for GLM-4.5.
- GPT-5.6 Luna accepts more context: 1.05M tokens versus 131K.
- GLM-4.5 has downloadable open weights; the other is API-only.
Side by side
| GLM-4.5 | GPT-5.6 Luna | |
|---|---|---|
| Provider | Z.ai (Zhipu) | OpenAI |
| Noometry Index | 41.4 | 54.7 |
| Released | 2025-07-27 | 2026-07-09 |
| Weights | Open | Proprietary |
| Context window | 131K | 1.05M |
| Max output | 98K | 128K |
| Input $ / M tokens | $0.60 | $0.20 |
| Output $ / M tokens | $2.20 | $1.20 |
| Results tracked | 27 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.6 Luna leads
GLM-4.5: 41.1 (#121), GPT-5.6 Luna: 53.9 (#25)
| Benchmark | GLM-4.5 | GPT-5.6 Luna |
|---|---|---|
| WeirdML | 40.6% | 60.9% |
| LMArena Coding | 1434 | 1466 |
| ALE-Bench | 344.82 | 1,667 |
| DeepSWE | — | 67.2% |
| FrontierCode | — | 39.8% |
| SWE-bench Verified (bash only) | 54.2% | — |
| CursorBench | — | 35.9% |
| LMArena WebDev | — | 1517 |
| SciCode | — | 53.6% |
| AlgoTune | 1.52 | — |
Agentic & Tool Use Not comparable
GLM-4.5: —, GPT-5.6 Luna: 35.3 (#37)
| Benchmark | GLM-4.5 | GPT-5.6 Luna |
|---|---|---|
| BALROG | — | 45.6% |
| GDP.pdf | — | 22.7% |
| Vending-Bench 2 | — | 4,095 |
Reasoning GPT-5.6 Luna leads
GLM-4.5: 27.7 (#99), GPT-5.6 Luna: 47.5 (#38)
| Benchmark | GLM-4.5 | GPT-5.6 Luna |
|---|---|---|
| Kagi LLM Benchmark | 57.9% | 49.1% |
| LMArena Hard Prompts | 1430 | 1452 |
| ARC-AGI-2 | — | 59.5% |
| SimpleBench | — | 46.8% |
| NYT Connections (extended) | — | 69.4% |
| ARC-AGI-1 | — | 88% |
| CritPt | — | 20.6% |
| Chess Puzzles | — | 40% |
| Mystery Game Puzzles | — | 21% |
| DTBench | — | 89.1% |
| LMCA | — | 48.5% |
| Surface Evolver Bench | — | 61.9% |
| Epoch Capabilities Index | — | 156.32 |
Math GPT-5.6 Luna leads
GLM-4.5: 36.3 (#119), GPT-5.6 Luna: 77.5 (#12)
| Benchmark | GLM-4.5 | GPT-5.6 Luna |
|---|---|---|
| LMArena Math | 1427 | 1455 |
| FrontierMath (Tiers 1-3) | — | 82.1% |
| FrontierMath Tier 4 | — | 61% |
| OTIS Mock AIME 2024-2025 | — | 98.3% |
| ProofBench | — | 60% |
Knowledge GPT-5.6 Luna leads
GLM-4.5: 35.5 (#172), GPT-5.6 Luna: 58.6 (#31)
| Benchmark | GLM-4.5 | GPT-5.6 Luna |
|---|---|---|
| LMArena Expert | 1433 | 1479 |
| GPQA Diamond | — | 91.6% |
| Humanity's Last Exam | 8.3% | — |
| SimpleQA Verified | — | 41% |
| Confabulations | 11.3% | — |
Multimodal Not comparable
GLM-4.5: —, GPT-5.6 Luna: 43.0 (#22)
| Benchmark | GLM-4.5 | GPT-5.6 Luna |
|---|---|---|
| LMArena Vision | — | 1259 |
| Blueprint-Bench 2 | — | 22.6% |
| Furniture Assembly | — | 42.5% |
| LMArena Document | — | 1457 |
Multilingual Too close to call
GLM-4.5: 53.0 (#74), GPT-5.6 Luna: 53.0 (#73)
| Benchmark | GLM-4.5 | GPT-5.6 Luna |
|---|---|---|
| LMArena Non-English | 1417 | 1418 |
| LMArena Chinese | 1465 | 1469 |
| LMArena French | 1423 | 1457 |
| LMArena German | 1406 | 1459 |
| LMArena Japanese | 1416 | 1416 |
| LMArena Korean | 1382 | 1414 |
| LMArena Russian | 1414 | 1428 |
| LMArena Spanish | 1454 | 1445 |
Instruction Following GPT-5.6 Luna leads
GLM-4.5: 74.3 (#100), GPT-5.6 Luna: 76.0 (#48)
| Benchmark | GLM-4.5 | GPT-5.6 Luna |
|---|---|---|
| LMArena Instruction Following | 1405 | 1440 |
Long Context GPT-5.6 Luna leads
GLM-4.5: 38.2 (#196), GPT-5.6 Luna: 43.9 (#78)
| Benchmark | GLM-4.5 | GPT-5.6 Luna |
|---|---|---|
| LMArena Longer Query | 1413 | 1437 |
| Fiction.LiveBench | 58.3% | — |
Writing & Preference GPT-5.6 Luna leads
GLM-4.5: 57.7 (#121), GPT-5.6 Luna: 68.3 (#27)
| Benchmark | GLM-4.5 | GPT-5.6 Luna |
|---|---|---|
| LMArena Text | 1430 | 1432 |
| LMArena Creative Writing | 1395 | 1397 |
| EQ-Bench Creative Writing | 1343 | 1829 |
| LMArena Multi-Turn | 1415 | 1433 |
| Short-Story Creative Writing | 73.4% | — |
| EQ-Bench 4 | — | 1156 |
Frequently asked questions
Is GLM-4.5 better than GPT-5.6 Luna?
GPT-5.6 Luna is the stronger model overall, scoring 54.7 to 41.4 on the Noometry Index.
Which is cheaper, GLM-4.5 or GPT-5.6 Luna?
GPT-5.6 Luna is cheaper. It lists at $0.20 per million input tokens and $1.20 per million output tokens; GLM-4.5 lists at $0.60 and $2.20.
Is GLM-4.5 or GPT-5.6 Luna better for coding?
GPT-5.6 Luna scores higher on coding benchmarks: 53.9 versus 41.1 in the Noometry coding category.
Which has the bigger context window?
GPT-5.6 Luna does, with 1.05M tokens against 131K.
How many benchmarks do GLM-4.5 and GPT-5.6 Luna share?
21 benchmarks have published results for both models. GLM-4.5 has 27 scored results on Noometry and GPT-5.6 Luna has 51.