Model comparison
GLM-5.3-Flash vs Step 3.5 Flash
GLM-5.3-Flash is the stronger model overall, scoring 51.8 to 42.3 on the Noometry Index. Step 3.5 Flash costs 1.6× less per token, which makes it the better buy when GLM-5.3-Flash's lead doesn't matter for your workload.
Last verified . 17 shared benchmarks.
StepFunSummary
- They share 17 benchmarks with published results for both. GLM-5.3-Flash scores higher in 8 categories and Step 3.5 Flash in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GLM-5.3-Flash leads 48.0 to 22.2.
- Step 3.5 Flash is cheaper at $0.10 / $0.30 per million input/output tokens, against $0.15 / $0.50 for GLM-5.3-Flash.
- GLM-5.3-Flash accepts more context: 1M tokens versus 256K.
Side by side
| GLM-5.3-Flash | Step 3.5 Flash | |
|---|---|---|
| Provider | Z.ai (Zhipu) | StepFun |
| Noometry Index | 51.8 | 42.3 |
| Released | 2026-08-20 | 2026-01-29 |
| Weights | Open | Open |
| Context window | 1M | 256K |
| Max output | 131K | 256K |
| Input $ / M tokens | $0.15 | $0.10 |
| Output $ / M tokens | $0.50 | $0.30 |
| Results tracked | 40 | 19 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.3-Flash leads
GLM-5.3-Flash: 53.1 (#31), Step 3.5 Flash: 42.4 (#105)
| Benchmark | GLM-5.3-Flash | Step 3.5 Flash |
|---|---|---|
| LMArena Coding | 1508 | 1436 |
| DeepSWE | 63.4% | — |
| FrontierCode | 31.8% | — |
| CursorBench | 36.8% | — |
| LMArena WebDev | 1609 | — |
| FrontierSWE | 18.1% | — |
| SciCode | 51.6% | — |
| ALE-Bench | 303.55 | — |
Agentic & Tool Use Not comparable
GLM-5.3-Flash: 34.2 (#47), Step 3.5 Flash: —
| Benchmark | GLM-5.3-Flash | Step 3.5 Flash |
|---|---|---|
| APEX-Agents | 52.8% | — |
| GDP.pdf | 14% | — |
Reasoning GLM-5.3-Flash leads
GLM-5.3-Flash: 48.0 (#42), Step 3.5 Flash: 22.2 (#202)
| Benchmark | GLM-5.3-Flash | Step 3.5 Flash |
|---|---|---|
| LMArena Hard Prompts | 1491 | 1411 |
| ARC-AGI-2 | 65.8% | — |
| NYT Connections (extended) | — | 28.4% |
| ARC-AGI-1 | 91% | — |
| CritPt | 15.4% | — |
| Chess Puzzles | 14% | — |
| Mystery Game Puzzles | 8% | — |
| Surface Evolver Bench | 52.5% | — |
| Bench to the Future 3 | 0.15 | — |
| Epoch Capabilities Index | 151.88 | — |
Math GLM-5.3-Flash leads
GLM-5.3-Flash: 53.3 (#47), Step 3.5 Flash: 42.6 (#84)
| Benchmark | GLM-5.3-Flash | Step 3.5 Flash |
|---|---|---|
| LMArena Math | 1500 | 1408 |
| FrontierMath (Tiers 1-3) | 55.8% | — |
| FrontierMath Tier 4 | 17.1% | — |
| MathArena Final-Answer Competitions | — | 66.8% |
| OTIS Mock AIME 2024-2025 | 93.9% | — |
| ProofBench | 21% | — |
Knowledge GLM-5.3-Flash leads
GLM-5.3-Flash: 58.4 (#36), Step 3.5 Flash: 39.6 (#132)
| Benchmark | GLM-5.3-Flash | Step 3.5 Flash |
|---|---|---|
| LMArena Expert | 1513 | 1421 |
| GPQA Diamond | 90.2% | — |
Multimodal Not comparable
GLM-5.3-Flash: 42.8 (#27), Step 3.5 Flash: —
| Benchmark | GLM-5.3-Flash | Step 3.5 Flash |
|---|---|---|
| LMArena Vision | 1296 | — |
Multilingual GLM-5.3-Flash leads
GLM-5.3-Flash: 56.0 (#25), Step 3.5 Flash: 50.5 (#119)
| Benchmark | GLM-5.3-Flash | Step 3.5 Flash |
|---|---|---|
| LMArena Non-English | 1462 | 1385 |
| LMArena Chinese | 1527 | 1447 |
| LMArena French | 1496 | 1421 |
| LMArena German | 1470 | 1405 |
| LMArena Japanese | 1429 | 1354 |
| LMArena Korean | 1446 | 1352 |
| LMArena Russian | 1469 | 1385 |
| LMArena Spanish | 1471 | 1419 |
Instruction Following GLM-5.3-Flash leads
GLM-5.3-Flash: 77.5 (#20), Step 3.5 Flash: 73.1 (#124)
| Benchmark | GLM-5.3-Flash | Step 3.5 Flash |
|---|---|---|
| LMArena Instruction Following | 1478 | 1385 |
Long Context GLM-5.3-Flash leads
GLM-5.3-Flash: 45.4 (#39), Step 3.5 Flash: 42.8 (#117)
| Benchmark | GLM-5.3-Flash | Step 3.5 Flash |
|---|---|---|
| LMArena Longer Query | 1482 | 1402 |
Writing & Preference GLM-5.3-Flash leads
GLM-5.3-Flash: 65.3 (#50), Step 3.5 Flash: 58.8 (#113)
| Benchmark | GLM-5.3-Flash | Step 3.5 Flash |
|---|---|---|
| LMArena Text | 1471 | 1403 |
| LMArena Creative Writing | 1442 | 1357 |
| LMArena Multi-Turn | 1467 | 1405 |
Frequently asked questions
Is GLM-5.3-Flash better than Step 3.5 Flash?
GLM-5.3-Flash is the stronger model overall, scoring 51.8 to 42.3 on the Noometry Index. Step 3.5 Flash costs 1.6× less per token, which makes it the better buy when GLM-5.3-Flash's lead doesn't matter for your workload.
Which is cheaper, GLM-5.3-Flash or Step 3.5 Flash?
Step 3.5 Flash is cheaper. It lists at $0.10 per million input tokens and $0.30 per million output tokens; GLM-5.3-Flash lists at $0.15 and $0.50.
Is GLM-5.3-Flash or Step 3.5 Flash better for coding?
GLM-5.3-Flash scores higher on coding benchmarks: 53.1 versus 42.4 in the Noometry coding category.
Which has the bigger context window?
GLM-5.3-Flash does, with 1M tokens against 256K.
How many benchmarks do GLM-5.3-Flash and Step 3.5 Flash share?
17 benchmarks have published results for both models. GLM-5.3-Flash has 40 scored results on Noometry and Step 3.5 Flash has 19.