Compare / Muse Spark 1.1 vs Muse Spark 1.2
Muse Spark 1.1 vs Muse Spark 1.2
22 results from 4 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | Muse Spark 1.1 | Muse Spark 1.2 |
|---|---|---|---|
| Arena (formerly LMArena)reported | Confirmed task successIPS effect estimate | -0.02 | -0.05 |
| Arena (formerly LMArena)reported | Praise over complaintIPS effect estimate | -0.11 | -0.12 |
| Arena (formerly LMArena)reported | SteerabilityIPS effect estimate | -0.06 | -0.05 |
| Arena (formerly LMArena)reported | Tool groundingIPS effect estimate | 0.00 | 0.00 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,463 | 1,484 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,493 | 1,519 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,449 | 1,448 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,504 | 1,502 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,473 | 1,475 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,491 | 1,496 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,460 | 1,472 |
| Artificial Analysisreported | Artificial Analysis Coding Indexindex score | 71.3 | 72.2 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 33.7 | 39.6 |
| Artificial Analysisreported | GPQA Diamond% of questions | 89.8% | 90.4% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 46.2% | 45.5% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 77.7% | 79.0% |
| Artificial Analysisreported | SciCode% of problems | 58.8% | 57.4% |
| Artificial Analysisreported | Terminal-Bench 2.1% of tasks | 77.9% | 80.2% |
| Artificial Analysisreported | Terminal-Bench 4.0% of tasks | 6.1% | 7.1% |
| Artificial Analysisreported | Τ-bench banking% of tasks | 31.8% | 34.8% |
| GDP.pdfreported | Professional document tasks% of rubric | 15.0% | 16.0% |
| SimpleQA Verified (Epoch AI)reported | Correct answers% of questions | 57.8% | 60.3% |
Bold green marks the better value on that metric. Where a source reports ranges that overlap, the difference may not be meaningful; see the benchmark page for ranges.