Muse Spark 1.2 vs Step 5 Preview
Muse Spark 1.2 
Step 5 Preview Verdict
- Muse Spark 1.2 (xhigh) and Step 5 Preview are level on quality (64.1 vs 65.0).
- Muse Spark 1.2 (xhigh) is stronger in agents & tools, coding, human preference, multimodal.
- Step 5 Preview is stronger in composite, knowledge, long context, reasoning.
- Step 5 Preview is 1.4× cheaper ($1.43 vs $2.00 per 1M blended).
| Metric | Muse Spark 1.2 (xhigh) | Step 5 Preview |
|---|---|---|
| BenchLeader Index | 64.1 | 65.0 |
| Agents & tools score | 57.5 | – |
| Coding score | 57.2 | – |
| Composite score | 69.9 | 84.4 |
| Human preference score | 70.6 | – |
| Knowledge score | 67.9 | 72.0 |
| Long context score | 65.9 | 70.8 |
| Multimodal score | 66.0 | 60.8 |
| Reasoning score | 73.8 | 82.1 |
| Blended price $/M | $2.00 | $1.43 |
| Output speed | 207 tok/s | – |
| Time to first answer | 22.3 s | – |
| Context window | 1.0M | 1M |
| SimpleQA Verified | 60.3% | – |
| SciCode | 56.4% | – |
| WeirdML | 60.3% | – |
| LMArena Text | 1500 | – |
| LMArena Hard Prompts | 1513 | – |
| LMArena Coding | 1536 | – |
| LMArena WebDev | 1534 | – |
| LMArena Vision | 1304 | – |
| LMArena Agent | -1.8 | – |
| LiveBench | 78.0% | – |
| LiveBench Reasoning | 90.0% | – |
| LiveBench Coding | 77.5% | – |
| LiveBench Agentic Coding | 57.6% | – |
| LiveBench Mathematics | 91.2% | – |
| LiveBench Data Analysis | 76.5% | – |
| LiveBench Language | 78.6% | – |
| LiveBench Instruction Following | 74.3% | – |
| AA Intelligence Index | 39.8 | 43.6 |
| AA-LCR | 79.0% | 88.3% |
| MMMU-Pro | – | 76.4% |
| AA-Omniscience | 27.2 | 16.4 |
| GPQA Diamond (AA) | 90.4% | – |
| Humanity's Last Exam (AA) | 45.5% | 46.5% |
| SciCode (AA) | 57.4% | 58.9% |
| MMLU-Pro | 88.3% | – |
| IOI | 21.8% | – |
| LegalBench | 85.3% | – |
| CorpFin | 70.9% | – |
| TaxEval | 80.4% | – |
| Terminal-Bench 2.1 (Vals) | 69.7% | – |
| SWE-bench (Vals) | 86.6% | – |
| Vals Index | 57.0 | – |
| CritPt | 17.7% | 20.9% |
| GDPval (AA) | 51.2% | 53.6% |
| τ²-Bench Banking (AA) | 34.9% | – |
| BioMysteryBench | 64.8% | – |
| Code Migration | 29.9% | – |
| Excel Modeling Benchmark | 57.0% | – |
| Finance Agent v2 | 60.6% | – |
| Harvey's Legal Agent Benchmark | 25.4% | – |
| Legal Research Bench | 43.8% | – |
| MedCode | 49.4% | – |
| MedScribe | 90.1% | – |
| MMMU-Pro (Vals) | 86.1% | – |
| MortgageTax | 65.4% | – |
| Public Benefits Bench | 68.5% | – |
| SAGE | 47.7% | – |
| SkillsBench | 53.0% | – |
| Terminal-Bench 4.0 (Vals) | 5.6% | – |
| Vibe Code Bench v1.1 | 79.1% | – |
| LMArena Creative Writing | 1452 | – |
| LMArena Instruction Following | 1480 | – |
| LMArena Multi-turn | 1520 | – |
| LMArena Longer Queries | 1502 | – |
| DeepSWE | 54.9% | – |
| LMCA | 48.4% | – |
| DTBench | 94.7% | – |
| GDP.pdf | 16.0% | – |
| FrontierSWE | 12.0% | – |
Data as of 2026-09-19. Best configuration of each model; every score links to its source on the model pages.
Muse Spark 1.2 vs Step 5 Preview: questions
- Is Muse Spark 1.2 better than Step 5 Preview?
- Muse Spark 1.2 (xhigh) and Step 5 Preview are level on quality (64.1 vs 65.0). The BenchLeader Index combines every independent quality benchmark; Step 5 Preview is ahead overall as of 2026-09-19, but check the category scores for your use.
- Which is cheaper, Muse Spark 1.2 or Step 5 Preview?
- Step 5 Preview is cheaper: $1.43 against $2.00 per million tokens, blended at three input tokens per output token.
- Which has the larger context window?
- Muse Spark 1.2 accepts more context: 1.0M against 1M tokens.