DeepSeek V4.1 Flash 1017 vs Muse Ember
TryTop 2 now
Muse Ember leads by 1.4 capability points (≈95% interval −1.2 to +4.0): statistically tied · P(higher) ≈ 86%
Wins 36 of 52 shared benchmarks (16 lost)
Capability vs. relative cost
DeepSeek V4.1 Flash 1017Muse EmberClosed weightsOpen weights
Benchmark wins16 – 36 · 52 shared
Comparing matching setups only (5 benchmarks excluded) · suites counted once
Area← DeepSeek V4.1 Flash 1017Muse Ember →
Data: Artificial Analysis, Vals AI, Arena and 6 othersatlas.village.dev
Margins
Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is Muse Ember minus DeepSeek V4.1 Flash 1017.
| Code Migration v1.1 (Vals Index v2.2 subset) | Software › Software engineering | 68.5 | 47.1 | −21.4 |
| RiemannBench (November 2026 rescoring) | Math | 38.4 | 53.6 | +15.2 |
| DAXBench v2 | Software › Data & ML | 77.7 | 62.8 | −14.9 |
| CritPt v2 (agentic, avg@5) | Science › Physical sciences | 50.6 | 62.3 | +11.7 |
| DAYJOB: Finance (October 2026 rescoring) | Professional › Finance, accounting & tax | 12.5 | 24.2 | +11.7 |
| τ³-Banking (Mercor run, bm25_grep retrieval) | Tool & computer use › Customer service & operations | 41.6 | 53.3 | +11.7 |
| DAYJOB: Healthcare (October 2026 rescoring) | Health & medicine | 6.0 | 17.2 | +11.2 |
| HLAB-Vals-1.1 | Professional › Legal | 14.2 | 25.0 | +10.8 |
| BuseyBench SVG | Perception & media › Generation | 4.3 | 5.4 | +10.5 |
| Ramp SWE-Bench v2 | Software › Software engineering | 60.8 | 51.8 | −9.0 |
| Medical Long Context Reasoning (MLCR-AA, October 2026 methodology) | Health & medicine | 26.7 | 18.3 | −8.3 |
| Terminal-Bench 4.0 | Software › Software engineering | 31.8 | 39.4 | +7.6 |
| Terminal-Bench-Science 0.1 | Science › Multidisciplinary | 12.9 | 20.0 | +7.1 |
| Terminal-Bench 4.0 (Artificial Analysis) | Software › Software engineering | 33.8 | 40.9 | +7.1 |
| LMArena Code Arena WebDev | Software › Frontend & design | 1659 | 1608 | −6.7 |
| AA-Briefcase v1.1 Elo | Professional › General knowledge work | 1416 | 1526 | +6.0 |
| HANDBOOK.md Agents v1.1 | Tool & computer use › Customer service & operations | 24.0 | 29.8 | +5.8 |
| GDP.pdf (Artificial Analysis, All-pass) | Professional › General knowledge work | 17.2 | 22.8 | +5.6 |
| Humanity's Last Exam Diamond | Knowledge & reasoning › General knowledge | 20.8 | 25.1 | +4.3 |
| Artificial Analysis Engineering Index | Science › Engineering | 37.6 | 41.8 | +4.2 |
| Vibe Code Bench v1.1 | Software › Frontend & design | 87.7 | 83.7 | −4.0 |
| Harvey LAB (Mercor run, undisclosed ~213-task cohort) | Professional › Legal | 18.8 | 22.7 | +3.9 |
| Legal Research Bench | Professional › Legal | 45.2 | 49.0 | +3.8 |
| Vals Legal Index | Professional › Legal | 71.9 | 75.7 | +3.8 |
| APEX-SWE | Software › Software engineering | 49.1 | 52.9 | +3.8 |
| APEX-Agents 1.1 | Professional › General knowledge work | 79.4 | 75.7 | −3.7 |
| EnterpriseBench: CoreCraft | Tool & computer use › Customer service & operations | 58.5 | 62.1 | +3.6 |
| Tax Agent Bench | Professional › Finance, accounting & tax | 74.9 | 71.4 | −3.6 |
| Artificial Analysis Intelligence Index v5.0.1 | Composite indexes | 32.6 | 36.0 | +3.4 |
| AA-Omniscience v1.1 Index | Knowledge & reasoning › General knowledge | 1.5 | 7.7 | +3.1 |
| GDP.xlsx | Professional › General knowledge work | 18.4 | 15.4 | −3.0 |
| Excel Modeling Benchmark v1.1 | Professional › Finance, accounting & tax | 69.2 | 72.2 | +3.0 |
| Artificial Analysis Coding Agent Index v2.1 | Software › Software engineering | 36.8 | 39.7 | +2.9 |
| LMArena Agent Arena Bash Reliability | Tool & computer use › General agents | 1.1 | 6.6 | +2.8 |
| Artificial Analysis Economics Index | Composite indexes | 38.5 | 41.0 | +2.5 |
| Artificial Analysis Finance & Accounting Index | Professional › Finance, accounting & tax | 42.2 | 44.6 | +2.4 |
| LMArena Agent Arena Confirmed Success | Tool & computer use › General agents | 2.2 | -2.0 | −2.1 |
| APEX-1 (Mercor, mean score) | Professional › General knowledge work | 68.3 | 70.4 | +2.1 |
| Artificial Analysis Progress Index v1 | Composite indexes | 324 | 329 | +1.8 |
| FrontierCode 2.0 (Artificial Analysis) | Software › Software engineering | 28.5 | 27.2 | −1.3 |
| LMArena Agent Arena Steering Burden | Tool & computer use › General agents | 2.3 | 4.8 | +1.2 |
| GBENCH One-shot Coding GScore | Games | 55.4 | 56.4 | +0.9 |
| Finance Agent v2.1 | Professional › Finance, accounting & tax | 62.9 | 63.8 | +0.9 |
| Vals Index v2.2 | Composite indexes | 61.3 | 60.5 | −0.8 |
| Artificial Analysis Legal Index | Professional › Legal | 37.7 | 38.5 | +0.7 |
| LMArena Agent Arena Praise vs Complaint | Tool & computer use › General agents | -1.5 | -2.7 | −0.6 |
| Artificial Analysis Strategy & Ops Index | Professional › General knowledge work | 49.2 | 48.7 | −0.5 |
| LMArena Agent Arena Tool Hallucination | Tool & computer use › Tool calling | -0.1 | 0.9 | +0.5 |
| AutomationBench-AA 1.0.6 | Tool & computer use › Customer service & operations | 68.5 | 68.1 | −0.4 |
| LMArena Agent Arena Net Improvement (Steering Burden) | Tool & computer use › General agents | 0.8 | 1.0 | +0.4 |
| Artificial Analysis Healthcare & Medical Index | Health & medicine | 41.3 | 41.3 | 0.0 |
| Vals Finance Index | Professional › Finance, accounting & tax | 60.9 | 60.9 | 0.0 |
Data: Artificial Analysis, Vals AI, Arena and 6 othersatlas.village.dev