Model comparison
Muse Spark 1.2 vs Step 3.5 Flash
Muse Spark 1.2 is the stronger model overall, scoring 50.3 to 42.3 on the Noometry Index. Step 3.5 Flash costs 13× less per token, which makes it the better buy when Muse Spark 1.2's lead doesn't matter for your workload.
Last verified . 15 shared benchmarks.
Meta
StepFunSummary
- They share 15 benchmarks with published results for both. Muse Spark 1.2 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 Muse Spark 1.2 leads 51.3 to 22.2.
- The biggest single-benchmark swing is NYT Connections (extended): 79.2% for Muse Spark 1.2 and 28.4% for Step 3.5 Flash.
- Step 3.5 Flash is cheaper at $0.10 / $0.30 per million input/output tokens, against $1.25 / $4.25 for Muse Spark 1.2.
- Muse Spark 1.2 accepts more context: 1.05M tokens versus 256K.
- Step 3.5 Flash has downloadable open weights; the other is API-only.
Side by side
| Muse Spark 1.2 | Step 3.5 Flash | |
|---|---|---|
| Provider | Meta | StepFun |
| Noometry Index | 50.3 | 42.3 |
| Released | 2026-08-05 | 2026-01-29 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 256K |
| Max output | 131K | 256K |
| Input $ / M tokens | $1.25 | $0.10 |
| Output $ / M tokens | $4.25 | $0.30 |
| Results tracked | 31 | 19 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Muse Spark 1.2 leads
Muse Spark 1.2: 49.2 (#51), Step 3.5 Flash: 42.4 (#105)
| Benchmark | Muse Spark 1.2 | Step 3.5 Flash |
|---|---|---|
| LMArena Coding | 1495 | 1436 |
| DeepSWE | 54.9% | — |
| LMArena WebDev | 1533 | — |
| FrontierSWE | 12% | — |
| SciCode | 56.4% | — |
| WeirdML | 60.3% | — |
Agentic & Tool Use Not comparable
Muse Spark 1.2: 29.4 (#87), Step 3.5 Flash: —
| Benchmark | Muse Spark 1.2 | Step 3.5 Flash |
|---|---|---|
| APEX-Agents | 36.4% | — |
| GDP.pdf | 16% | — |
Reasoning Muse Spark 1.2 leads
Muse Spark 1.2: 51.3 (#34), Step 3.5 Flash: 22.2 (#202)
| Benchmark | Muse Spark 1.2 | Step 3.5 Flash |
|---|---|---|
| NYT Connections (extended) | 79.2% | 28.4% |
| LMArena Hard Prompts | 1486 | 1411 |
| SimpleBench | 74.5% | — |
| CritPt | 17.7% | — |
| DTBench | 94.7% | — |
| LMCA | 48.4% | — |
| Epoch Capabilities Index | 154.87 | — |
Math Muse Spark 1.2 leads
Muse Spark 1.2: 46.4 (#70), Step 3.5 Flash: 42.6 (#84)
| Benchmark | Muse Spark 1.2 | Step 3.5 Flash |
|---|---|---|
| LMArena Math | 1471 | 1408 |
| MathArena Final-Answer Competitions | — | 66.8% |
| ProofBench | 43% | — |
Knowledge Muse Spark 1.2 leads
Muse Spark 1.2: 54.1 (#53), Step 3.5 Flash: 39.6 (#132)
| Benchmark | Muse Spark 1.2 | Step 3.5 Flash |
|---|---|---|
| LMArena Expert | 1480 | 1421 |
| SimpleQA Verified | 60.3% | — |
Multimodal Not comparable
Muse Spark 1.2: 43.4 (#25), Step 3.5 Flash: —
| Benchmark | Muse Spark 1.2 | Step 3.5 Flash |
|---|---|---|
| LMArena Vision | 1305 | — |
Multilingual Muse Spark 1.2 leads
Muse Spark 1.2: 57.1 (#11), Step 3.5 Flash: 50.5 (#119)
| Benchmark | Muse Spark 1.2 | Step 3.5 Flash |
|---|---|---|
| LMArena Non-English | 1478 | 1385 |
| LMArena Chinese | 1511 | 1447 |
| LMArena French | 1513 | 1421 |
| LMArena Russian | 1487 | 1385 |
| LMArena Spanish | 1498 | 1419 |
| LMArena German | — | 1405 |
| LMArena Japanese | — | 1354 |
| LMArena Korean | — | 1352 |
Instruction Following Muse Spark 1.2 leads
Muse Spark 1.2: 76.7 (#36), Step 3.5 Flash: 73.1 (#124)
| Benchmark | Muse Spark 1.2 | Step 3.5 Flash |
|---|---|---|
| LMArena Instruction Following | 1461 | 1385 |
Long Context Muse Spark 1.2 leads
Muse Spark 1.2: 45.2 (#48), Step 3.5 Flash: 42.8 (#117)
| Benchmark | Muse Spark 1.2 | Step 3.5 Flash |
|---|---|---|
| LMArena Longer Query | 1475 | 1402 |
Writing & Preference Muse Spark 1.2 leads
Muse Spark 1.2: 72.3 (#14), Step 3.5 Flash: 58.8 (#113)
| Benchmark | Muse Spark 1.2 | Step 3.5 Flash |
|---|---|---|
| LMArena Text | 1482 | 1403 |
| LMArena Creative Writing | 1449 | 1357 |
| LMArena Multi-Turn | 1494 | 1405 |
| EQ-Bench Creative Writing | 1840 | — |
Frequently asked questions
Is Muse Spark 1.2 better than Step 3.5 Flash?
Muse Spark 1.2 is the stronger model overall, scoring 50.3 to 42.3 on the Noometry Index. Step 3.5 Flash costs 13× less per token, which makes it the better buy when Muse Spark 1.2's lead doesn't matter for your workload.
Which is cheaper, Muse Spark 1.2 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; Muse Spark 1.2 lists at $1.25 and $4.25.
Is Muse Spark 1.2 or Step 3.5 Flash better for coding?
Muse Spark 1.2 scores higher on coding benchmarks: 49.2 versus 42.4 in the Noometry coding category.
Which has the bigger context window?
Muse Spark 1.2 does, with 1.05M tokens against 256K.
How many benchmarks do Muse Spark 1.2 and Step 3.5 Flash share?
15 benchmarks have published results for both models. Muse Spark 1.2 has 31 scored results on Noometry and Step 3.5 Flash has 19.