Model comparison
GLM-5.3 vs GPT-5.6 Luna
GLM-5.3 and GPT-5.6 Luna score almost the same on the Noometry Index (54.8 vs 54.6), so choose on price, context window or the category you care about most.
Last verified . 40 shared benchmarks.
Summary
- They share 40 benchmarks with published results for both. GLM-5.3 scores higher in 6 categories and GPT-5.6 Luna in 3 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Luna leads 77.7 to 62.3.
- The biggest single-benchmark swing is FrontierMath Tier 4: 29.3% for GLM-5.3 and 61% for GPT-5.6 Luna.
- GPT-5.6 Luna is cheaper at $0.20 / $1.20 per million input/output tokens, against $1.40 / $4.40 for GLM-5.3.
- GPT-5.6 Luna accepts more context: 1.05M tokens versus 1M.
- GLM-5.3 has downloadable open weights; the other is API-only.
Side by side
| GLM-5.3 | GPT-5.6 Luna | |
|---|---|---|
| Provider | Z.ai (Zhipu) | OpenAI |
| Noometry Index | 54.8 | 54.6 |
| Released | 2026-08-14 | 2026-07-09 |
| Weights | Open | Proprietary |
| Context window | 1M | 1.05M |
| Max output | 131K | 128K |
| Input $ / M tokens | $1.40 | $0.20 |
| Output $ / M tokens | $4.40 | $1.20 |
| Results tracked | 42 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.3 leads
GLM-5.3: 59.5 (#14), GPT-5.6 Luna: 54.5 (#28)
| Benchmark | GLM-5.3 | GPT-5.6 Luna |
|---|---|---|
| DeepSWE | 69% | 67.2% |
| FrontierCode | 40.1% | 39.8% |
| CursorBench | 42.6% | 35.9% |
| LMArena WebDev | 1622 | 1519 |
| SciCode | 59% | 53.6% |
| WeirdML | 75.4% | 60.9% |
| LMArena Coding | 1496 | 1466 |
| ALE-Bench | 1,317 | 1,667 |
| FrontierSWE | 30.2% | — |
Agentic & Tool Use GLM-5.3 leads
GLM-5.3: 36.4 (#38), GPT-5.6 Luna: 34.4 (#45)
| Benchmark | GLM-5.3 | GPT-5.6 Luna |
|---|---|---|
| APEX-Agents | 56.6% | 43% |
| Vending-Bench 2 | 8,164 | 4,095 |
| BALROG | — | 45.6% |
| GDP.pdf | — | 22.7% |
Reasoning GPT-5.6 Luna leads
GLM-5.3: 46.1 (#46), GPT-5.6 Luna: 47.6 (#43)
| Benchmark | GLM-5.3 | GPT-5.6 Luna |
|---|---|---|
| NYT Connections (extended) | 74.2% | 69.4% |
| CritPt | 19.1% | 20.6% |
| Chess Puzzles | 21% | 40% |
| LMArena Hard Prompts | 1489 | 1451 |
| Mystery Game Puzzles | 33% | 21% |
| DTBench | 87.7% | 89.1% |
| LMCA | 55.5% | 48.5% |
| Epoch Capabilities Index | 155.61 | 156.39 |
| ARC-AGI-2 | — | 59.5% |
| SimpleBench | — | 46.8% |
| Kagi LLM Benchmark | — | 49.1% |
| ARC-AGI-1 | — | 88% |
| Surface Evolver Bench | — | 61.9% |
| Bench to the Future 3 | 0.15 | — |
Math GPT-5.6 Luna leads
GLM-5.3: 62.3 (#33), GPT-5.6 Luna: 77.7 (#14)
| Benchmark | GLM-5.3 | GPT-5.6 Luna |
|---|---|---|
| FrontierMath (Tiers 1-3) | 68.8% | 82.1% |
| FrontierMath Tier 4 | 29.3% | 61% |
| OTIS Mock AIME 2024-2025 | 91.1% | 98.3% |
| ProofBench | 49% | 60% |
| LMArena Math | 1489 | 1458 |
Knowledge Too close to call
GLM-5.3: 58.3 (#37), GPT-5.6 Luna: 58.5 (#34)
| Benchmark | GLM-5.3 | GPT-5.6 Luna |
|---|---|---|
| GPQA Diamond | 90.9% | 91.6% |
| SimpleQA Verified | 41% | 41% |
| LMArena Expert | 1516 | 1478 |
Multimodal Not comparable
GLM-5.3: —, GPT-5.6 Luna: 42.7 (#28)
| Benchmark | GLM-5.3 | GPT-5.6 Luna |
|---|---|---|
| LMArena Vision | — | 1258 |
| Blueprint-Bench 2 | — | 22.6% |
| Furniture Assembly | — | 42.5% |
| LMArena Document | — | 1457 |
Multilingual GLM-5.3 leads
GLM-5.3: 55.7 (#28), GPT-5.6 Luna: 52.8 (#78)
| Benchmark | GLM-5.3 | GPT-5.6 Luna |
|---|---|---|
| LMArena Non-English | 1457 | 1417 |
| LMArena Chinese | 1528 | 1470 |
| LMArena French | 1499 | 1456 |
| LMArena German | 1499 | 1454 |
| LMArena Japanese | 1453 | 1411 |
| LMArena Korean | 1472 | 1415 |
| LMArena Russian | 1463 | 1428 |
| LMArena Spanish | 1460 | 1448 |
Instruction Following GLM-5.3 leads
GLM-5.3: 77.5 (#23), GPT-5.6 Luna: 75.6 (#57)
| Benchmark | GLM-5.3 | GPT-5.6 Luna |
|---|---|---|
| LMArena Instruction Following | 1477 | 1437 |
Long Context GLM-5.3 leads
GLM-5.3: 45.4 (#41), GPT-5.6 Luna: 43.9 (#82)
| Benchmark | GLM-5.3 | GPT-5.6 Luna |
|---|---|---|
| LMArena Longer Query | 1482 | 1436 |
Writing & Preference GLM-5.3 leads
GLM-5.3: 75.7 (#6), GPT-5.6 Luna: 68.0 (#29)
| Benchmark | GLM-5.3 | GPT-5.6 Luna |
|---|---|---|
| LMArena Text | 1471 | 1431 |
| LMArena Creative Writing | 1457 | 1396 |
| EQ-Bench Creative Writing | 2075 | 1829 |
| LMArena Multi-Turn | 1472 | 1434 |
| EQ-Bench 4 | — | 1156 |
Frequently asked questions
Is GLM-5.3 better than GPT-5.6 Luna?
GLM-5.3 and GPT-5.6 Luna score almost the same on the Noometry Index (54.8 vs 54.6), so choose on price, context window or the category you care about most.
Which is cheaper, GLM-5.3 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-5.3 lists at $1.40 and $4.40.
Is GLM-5.3 or GPT-5.6 Luna better for coding?
GLM-5.3 scores higher on coding benchmarks: 59.5 versus 54.5 in the Noometry coding category.
Which has the bigger context window?
GPT-5.6 Luna does, with 1.05M tokens against 1M.
How many benchmarks do GLM-5.3 and GPT-5.6 Luna share?
40 benchmarks have published results for both models. GLM-5.3 has 42 scored results on Noometry and GPT-5.6 Luna has 52.