Model comparison
GLM-4.5-Air vs Step 3
Step 3 is the stronger model overall, scoring 40.5 to 38.9 on the Noometry Index.
Last verified . 16 shared benchmarks.
StepFunSummary
- They share 16 benchmarks with published results for both. GLM-4.5-Air scores higher in 3 categories and Step 3 in 5 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in coding, where Step 3 leads 40.1 to 33.3.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 43% for GLM-4.5-Air and 62.3% for Step 3.
Side by side
| GLM-4.5-Air | Step 3 | |
|---|---|---|
| Provider | Z.ai (Zhipu) | StepFun |
| Noometry Index | 38.9 | 40.5 |
| Released | 2025-07-20 | — |
| Weights | Open | Open |
| Context window | 131K | — |
| Max output | 98K | — |
| Input $ / M tokens | $0.20 | — |
| Output $ / M tokens | $1.10 | — |
| Results tracked | 27 | 17 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Step 3 leads
GLM-4.5-Air: 33.3 (#259), Step 3: 40.1 (#147)
| Benchmark | GLM-4.5-Air | Step 3 |
|---|---|---|
| LMArena Coding | 1397 | 1367 |
| GSO | 2.9% | — |
Reasoning Step 3 leads
GLM-4.5-Air: 24.1 (#166), Step 3: 28.4 (#105)
| Benchmark | GLM-4.5-Air | Step 3 |
|---|---|---|
| Kagi LLM Benchmark | 43% | 62.3% |
| LMArena Hard Prompts | 1379 | 1355 |
| ForecastBench | 59.2 | — |
Math Step 3 leads
GLM-4.5-Air: 36.2 (#170), Step 3: 37.6 (#148)
| Benchmark | GLM-4.5-Air | Step 3 |
|---|---|---|
| LMArena Math | 1396 | 1366 |
| Omni-MATH | 39.1% | — |
Knowledge Step 3 leads
GLM-4.5-Air: 35.0 (#191), Step 3: 36.8 (#164)
| Benchmark | GLM-4.5-Air | Step 3 |
|---|---|---|
| LMArena Expert | 1370 | 1333 |
| Humanity's Last Exam | 8.1% | — |
| MMLU-Pro | 76.2% | — |
| Vectara Hallucination Rate | 9.3% | — |
| GPQA (HELM) | 59.4% | — |
Multimodal Not comparable
GLM-4.5-Air: —, Step 3: 35.5 (#86)
| Benchmark | GLM-4.5-Air | Step 3 |
|---|---|---|
| LMArena Vision | — | 1177 |
Multilingual GLM-4.5-Air leads
GLM-4.5-Air: 49.1 (#135), Step 3: 46.3 (#159)
| Benchmark | GLM-4.5-Air | Step 3 |
|---|---|---|
| LMArena Non-English | 1366 | 1327 |
| LMArena Chinese | 1426 | 1397 |
| LMArena German | 1377 | 1371 |
| LMArena Korean | 1308 | 1269 |
| LMArena Russian | 1373 | 1331 |
| LMArena Spanish | 1386 | 1371 |
| LMArena French | 1399 | — |
| LMArena Japanese | 1348 | — |
Instruction Following Too close to call
GLM-4.5-Air: 69.6 (#171), Step 3: 70.4 (#164)
| Benchmark | GLM-4.5-Air | Step 3 |
|---|---|---|
| LMArena Instruction Following | 1354 | 1332 |
| IFEval | 81.2% | — |
Long Context GLM-4.5-Air leads
GLM-4.5-Air: 41.6 (#135), Step 3: 40.3 (#157)
| Benchmark | GLM-4.5-Air | Step 3 |
|---|---|---|
| LMArena Longer Query | 1366 | 1326 |
Writing & Preference GLM-4.5-Air leads
GLM-4.5-Air: 55.9 (#139), Step 3: 54.3 (#151)
| Benchmark | GLM-4.5-Air | Step 3 |
|---|---|---|
| LMArena Text | 1384 | 1350 |
| LMArena Creative Writing | 1343 | 1321 |
| LMArena Multi-Turn | 1371 | 1341 |
| WildBench | 78.9% | — |
Frequently asked questions
Is GLM-4.5-Air better than Step 3?
Step 3 is the stronger model overall, scoring 40.5 to 38.9 on the Noometry Index.
Is GLM-4.5-Air or Step 3 better for coding?
Step 3 scores higher on coding benchmarks: 40.1 versus 33.3 in the Noometry coding category.
How many benchmarks do GLM-4.5-Air and Step 3 share?
16 benchmarks have published results for both models. GLM-4.5-Air has 27 scored results on Noometry and Step 3 has 17.