Model comparison
GLM-5.2 vs Step 5 Preview
GLM-5.2 is the stronger model overall, scoring 51.1 to 47.9 on the Noometry Index. Step 5 Preview costs 1.5× less per token, which makes it the better buy when GLM-5.2's lead doesn't matter for your workload.
Last verified . 17 shared benchmarks.
StepFunSummary
- They share 17 benchmarks with published results for both. GLM-5.2 scores higher in 7 categories and Step 5 Preview in 1 category; 5 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GLM-5.2 leads 57.1 to 41.2.
- The biggest single-benchmark swing is SciCode: 50.5% for GLM-5.2 and 58.9% for Step 5 Preview.
- Step 5 Preview is cheaper at $1 / $2.70 per million input/output tokens, against $1.40 / $4.40 for GLM-5.2.
- Step 5 Preview accepts more context: 1.02M tokens versus 1M.
- GLM-5.2 has downloadable open weights; the other is API-only.
Side by side
| GLM-5.2 | Step 5 Preview | |
|---|---|---|
| Provider | Z.ai (Zhipu) | StepFun |
| Noometry Index | 51.1 | 47.9 |
| Released | 2026-06-13 | 2026-09-16 |
| Weights | Open | Proprietary |
| Context window | 1M | 1.02M |
| Max output | 131K | 66K |
| Input $ / M tokens | $1.40 | $1 |
| Output $ / M tokens | $4.40 | $2.70 |
| Results tracked | 51 | 18 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
GLM-5.2: 51.3 (#41), Step 5 Preview: 51.8 (#36)
| Benchmark | GLM-5.2 | Step 5 Preview |
|---|---|---|
| LMArena WebDev | 1603 | 1564 |
| SciCode | 50.5% | 58.9% |
| LMArena Coding | 1485 | 1480 |
| SWE-bench Verified | 78.7% | — |
| DeepSWE | 43.8% | — |
| FrontierCode | 24.5% | — |
| WeirdML | 70.1% | — |
| ALE-Bench | 1,047 | — |
Agentic & Tool Use Not comparable
GLM-5.2: 32.4 (#63), Step 5 Preview: —
| Benchmark | GLM-5.2 | Step 5 Preview |
|---|---|---|
| APEX-Agents | 45.2% | — |
| τ²-bench Banking | 37.1% | — |
| PostTrainBench | 31.7% | — |
| GBAEval | 0% | — |
| Vending-Bench 2 | 8,314 | — |
Reasoning GLM-5.2 leads
GLM-5.2: 42.3 (#52), Step 5 Preview: 40.0 (#57)
| Benchmark | GLM-5.2 | Step 5 Preview |
|---|---|---|
| CritPt | 20.9% | 20.9% |
| LMArena Hard Prompts | 1480 | 1465 |
| ARC-AGI-2 | 22.8% | — |
| SimpleBench | 58.8% | — |
| Kagi LLM Benchmark | 62.6% | — |
| NYT Connections (extended) | 74.3% | — |
| ARC-AGI-1 | 77% | — |
| Chess Puzzles | 21% | — |
| EBR-Bench | 9.5% | — |
| Mystery Game Puzzles | 19% | — |
| DTBench | 93.6% | — |
| LMCA | 45.8% | — |
| Surface Evolver Bench | 55.6% | — |
| Epoch Capabilities Index | 151.78 | — |
Math GLM-5.2 leads
GLM-5.2: 55.7 (#43), Step 5 Preview: 46.1 (#72)
| Benchmark | GLM-5.2 | Step 5 Preview |
|---|---|---|
| ProofBench | 35% | 42% |
| LMArena Math | 1482 | 1470 |
| FrontierMath (Tiers 1-3) | 59.2% | — |
| FrontierMath Tier 4 | 29.3% | — |
| MathArena Final-Answer Competitions | 67.6% | — |
| OTIS Mock AIME 2024-2025 | 86.4% | — |
Knowledge GLM-5.2 leads
GLM-5.2: 57.1 (#40), Step 5 Preview: 41.2 (#112)
| Benchmark | GLM-5.2 | Step 5 Preview |
|---|---|---|
| LMArena Expert | 1486 | 1470 |
| GPQA Diamond | 91.9% | — |
| SimpleQA Verified | 34.2% | — |
Multimodal Not comparable
GLM-5.2: —, Step 5 Preview: 41.0 (#41)
| Benchmark | GLM-5.2 | Step 5 Preview |
|---|---|---|
| LMArena Vision | — | 1267 |
Multilingual GLM-5.2 leads
GLM-5.2: 55.8 (#26), Step 5 Preview: 53.8 (#54)
| Benchmark | GLM-5.2 | Step 5 Preview |
|---|---|---|
| LMArena Non-English | 1459 | 1432 |
| LMArena Chinese | 1519 | 1519 |
| LMArena Russian | 1466 | 1436 |
| LMArena Spanish | 1477 | 1439 |
| LMArena French | 1479 | — |
| LMArena German | 1468 | — |
| LMArena Japanese | 1451 | — |
| LMArena Korean | 1445 | — |
Instruction Following Too close to call
GLM-5.2: 76.9 (#34), Step 5 Preview: 76.0 (#50)
| Benchmark | GLM-5.2 | Step 5 Preview |
|---|---|---|
| LMArena Instruction Following | 1465 | 1444 |
Long Context Too close to call
GLM-5.2: 45.3 (#43), Step 5 Preview: 44.5 (#64)
| Benchmark | GLM-5.2 | Step 5 Preview |
|---|---|---|
| LMArena Longer Query | 1479 | 1455 |
Writing & Preference GLM-5.2 leads
GLM-5.2: 70.4 (#21), Step 5 Preview: 62.8 (#71)
| Benchmark | GLM-5.2 | Step 5 Preview |
|---|---|---|
| LMArena Text | 1470 | 1442 |
| LMArena Creative Writing | 1462 | 1410 |
| LMArena Multi-Turn | 1469 | 1447 |
| EQ-Bench Creative Writing | 1757 | — |
| EQ-Bench 4 | 1222 | — |
Frequently asked questions
Is GLM-5.2 better than Step 5 Preview?
GLM-5.2 is the stronger model overall, scoring 51.1 to 47.9 on the Noometry Index. Step 5 Preview costs 1.5× less per token, which makes it the better buy when GLM-5.2's lead doesn't matter for your workload.
Which is cheaper, GLM-5.2 or Step 5 Preview?
Step 5 Preview is cheaper. It lists at $1 per million input tokens and $2.70 per million output tokens; GLM-5.2 lists at $1.40 and $4.40.
Is GLM-5.2 or Step 5 Preview better for coding?
They score almost the same on coding (51.3 vs 51.8); test both on your own repository before choosing.
Which has the bigger context window?
Step 5 Preview does, with 1.02M tokens against 1M.
How many benchmarks do GLM-5.2 and Step 5 Preview share?
17 benchmarks have published results for both models. GLM-5.2 has 51 scored results on Noometry and Step 5 Preview has 18.