Muse Ember vs Qwen3.5-27B
Full profiles: Muse Ember (Meta) and Qwen3.5-27B (Alibaba): every benchmark, price and rank.
Updated · Data from 7 sources
As of 16 Nov 2026, there are not enough shared benchmarks for a verdict, and Muse Ember is 4.1× cheaper at $0.200 vs $0.825 per 1M tokens blended. Choose Muse Ember for coding (60 vs 44), where it also streams 3×+ faster; choose Qwen3.5-27B for fast first answers (2.9 s vs 5.0 s) and reasoning (58 vs 51).
Key facts (as of 16 Nov 2026)
- Muse Ember scores 52.1 on the BenchLeader Index and Qwen3.5-27B scores 52.0 (BenchLeader, 16 Nov 2026).
- Muse Ember costs $0.200 per 1M tokens blended (list price, 16 Nov 2026).
- Qwen3.5-27B costs $0.825 per 1M tokens blended (list price, 16 Nov 2026).
- Muse Ember streams 3×+ faster than Qwen3.5-27B (Artificial Analysis, 7-day median, 16 Nov 2026).
- Of the 1 benchmarks both models have, Muse Ember leads 1 (BenchLeader, 16 Nov 2026).
- On Arena WebDev, Muse Ember is rated 1609 and Qwen3.5-27B 1357 (LMArena, as published by 16 Nov 2026).
Which to pick for…
The pick's numbers come first. Category scores on the BenchLeader scale (50 is the average model, 15 points one standard deviation). High confidence means a lead of half a standard deviation or more; under one point is called either.
| Use case | Pick | Why | Confidence |
|---|---|---|---|
| Coding | Muse Ember | 60 vs 44 category score; Arena WebDev 1609 vs 1357 | High |
| Reasoning | Qwen3.5-27B | 58 vs 51 category score | Medium |
| Chat quality (human preference) | – | Only Qwen3.5-27B has an Arena rating (58 category score) | Not comparable |
| Lowest cost | Muse Ember | $0.200 vs $0.825 per 1M blended; agentic coding session $0.008 vs $0.028 | – |
| Fastest first answer | Qwen3.5-27B | 2.9 s (Qwen3.5-27B) vs 5.0 s (Muse Ember (max)) to the first answer token | – |
| High volume on a budget | Neither flagship | A smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs | – |
Key differences
- Not enough shared benchmarks for a verdict.
- Muse Ember (max) is stronger in coding.
- Qwen3.5-27B is stronger in reasoning.
- Muse Ember (max) is 4.1× cheaper ($0.200 vs $0.825 per 1M blended).
- Muse Ember (max) streams 11.6× faster (437 vs 38 tokens per second).
At a glance
| Spec | Muse Ember | Qwen3.5-27B |
|---|---|---|
| Developer | Meta | Alibaba |
| Released | 9 Nov 2026 | 24 Feb 2026 |
| Context window (tokens) | 1,048,576 | 262,144 |
| Compared at | Muse Ember (max) | Qwen3.5-27B |
| Open weights | No | Yes |
| Price in / out / cached input per 1M | $0.100 / $0.500 / $0.0040 | $0.300 / $2.40 / – |
Category by category
Category scores on the BenchLeader scale: 50 is the average evaluated model, 15 points is one standard deviation. A gap under one point is called level.
| Category | Muse Ember (max) | Qwen3.5-27B | Leader |
|---|---|---|---|
| Reasoning | 51.1 | 58.3 | Qwen3.5-27B |
| Coding | 60.0 | 44.3 | Muse Ember |
| Agents & tools | 52.4 | – | not comparable |
| Maths | – | 45.6 | not comparable |
| Knowledge | 48.8 | – | not comparable |
| Human preference | – | 58.2 | not comparable |
| Multimodal | – | 56.8 | not comparable |
On one chart
Each domain on the index's own scale, where 50 is an average model. Click a domain to see the benchmarks behind it. Pick up to 5 models from the list.
Dashed: that model has no result on the axis it crosses; the line only joins its neighbours.
Not drawn, no result for these models: Instruction following, Long context.
Where sources disagree
Across the 1 benchmarks both have, Muse Ember leads 1, Qwen3.5-27B leads 0.
| Overall index | Muse Ember (max) | Qwen3.5-27B | Leader |
|---|---|---|---|
| BenchLeader Index | 52.1 | 52.0 | neck and neck |
Each publisher runs its own tasks, harnesses and effort settings, so their overall verdicts can differ; BenchLeader reconciles them on one scale. † measured at another effort setting of the same model (the publisher did not test the one compared here).
Price, speed and latency
List prices blended at three input tokens per output token. Per-task prices use the token counts on the methodology page. Cost, tokens and time per benchmark task are measured: Artificial Analysis’ averages over its whole Intelligence Index run.
| Measure | Muse Ember (max) | Qwen3.5-27B |
|---|---|---|
| Blended price per 1M tokens | $0.200 | $0.825 |
| Input / output per 1M | $0.100 / $0.500 | $0.300 / $2.40 |
| Cached input per 1M | $0.0040 | – |
| Price per task: chat reply | $0.0002 | $0.0008 |
| Price per task: summarize a 30-page report | $0.0015 | $0.0050 |
| Price per task: code edit | $0.0014 | $0.0054 |
| Price per task: agentic coding session | $0.0080 | $0.028 |
| Price per task: structured extraction | $0.0003 | $0.0011 |
| Agentic coding session, input mostly cached | $0.0037 | – |
| Cost per benchmark task (AA Index) | $0.136 | – |
| Output tokens per task (AA Index) | 84k | – |
| Of which reasoning | 52k | – |
| Time per task (AA Index) | 3 min | – |
| Output speed | 437 tok/s | 38 tok/s |
| Time to first answer | 5.0 s | 2.9 s |
| Time to a complete answer, last measured 13 Nov 2026 | 12.9 s | – |
Each model is compared at its best-ranked reasoning-effort setting. Muse Ember (max) scores 52.1 and takes 5 s to its first answer token; Muse Ember (xhigh) scores 49.1 in 4 s. Qwen3.5-27B scores 52.0 and takes 3 s to its first answer token. Higher effort means more thinking before the answer: better scores, longer waits and more tokens billed. How reasoning effort changes scores and cost.
Every benchmark
66 benchmarks, grouped by category; each name links to the board and its leaders. † measured at another effort setting of the same model, shown for reference and not counted as a win.
| Metric | Publisher | Muse Ember (max) | Qwen3.5-27B |
|---|---|---|---|
| BenchLeader Index | BenchLeader | 52.1 | 52.0 |
| Agents & tools score | BenchLeader | 52.4 | – |
| Coding score | BenchLeader | 60.0 | 44.3 |
| Knowledge score | BenchLeader | 48.8 | – |
| Reasoning score | BenchLeader | 51.1 | 58.3 |
| Human preference score | BenchLeader | – | 58.2 |
| Maths score | BenchLeader | – | 45.6 |
| Multimodal score | BenchLeader | – | 56.8 |
| Blended price $/M | BenchLeader | $0.200 | $0.825 |
| Output speed | BenchLeader | 437 tok/s | 38 tok/s |
| Time to first answer | BenchLeader | 5.0 s | 2.9 s |
| Context window | BenchLeader | 1,048,576 tokens | 262,144 tokens |
| Reasoning | |||
| SimpleBench | SimpleBench | 72.7%† | – |
| Arena Hard Prompts | Arena (LMArena) | – | 1427 |
| GPQA Diamond (AA) | Artificial Analysis | – | 84.2%† |
| Humanity's Last Exam (AA) | Artificial Analysis | – | 13.9%† |
| HLE-Diamond (AA) | Artificial Analysis | 25.1% | – |
| CritPt | Artificial Analysis | – | 0.3%† |
| CritPt v2 hard subset (AA) | Artificial Analysis | 39.0% | – |
| Arena Expert | Arena (LMArena) | – | 1441 |
| LMCA | Epoch AI Benchmarking Hub | – | 34.0% |
| DTBench | Epoch AI Benchmarking Hub | – | 82.4% |
| Coding | |||
| WeirdML | WeirdML | – | 39.5% |
| Arena Coding | Arena (LMArena) | – | 1452 |
| Arena WebDev | Arena (LMArena) | 1609 | 1357 |
| Vibe Code Bench v1.1 | Vals AI | 83.7% | – |
| SWE Atlas v1.1 | Scale AI SEAL | 56.0%† | – |
| SWE Atlas v1.1: Codebase QnA | Scale AI SEAL | 61.3%† | – |
| SWE Atlas v1.1: Test Writing | Scale AI SEAL | 56.7%† | – |
| SWE Atlas v1.1: Refactoring | Scale AI SEAL | 50.0%† | – |
| ALE-Bench | Epoch AI Benchmarking Hub | – | 349.4 |
| Terminal-Bench 4.0 (AA) | Artificial Analysis | 40.9% | – |
| Design Arena Website | Design Arena | 1330† | – |
| FrontierCode 2.0 (AA) | Artificial Analysis | 27.2% | – |
| Agents & tools | |||
| Terminal-Bench | Terminal-Bench | 30.6%† | – |
| APEX-Agents | Mercor | 64.3%† | – |
| Arena Agent | Arena (LMArena) | 1.7 | – |
Show all 66 benchmarks
| Metric | Publisher | Muse Ember (max) | Qwen3.5-27B |
|---|---|---|---|
| Agents & tools | |||
| Terminal-Bench Hard | Artificial Analysis | – | 31.8%† |
| τ²-Bench Telecom (AA) | Artificial Analysis | – | 87.1%† |
| ITBench SRE (AA) | Artificial Analysis | – | 35.5%† |
| Finance Agent v2.1 | Vals AI | 63.8% | – |
| Harvey's Legal Agent Benchmark v1.1 | Vals AI | 25.0% | – |
| Legal Research Bench | Vals AI | 49.0% | – |
| Tax Agent Bench | Vals AI | 38.2% | – |
| Terminal-Bench 4.0 (Vals) | Vals AI | 39.4% | – |
| Arena Agent Task Completion | Arena (LMArena) | -7.6 | – |
| Vending-Bench 2 | Andon Labs | – | 202.0 |
| AutomationBench | Artificial Analysis | 68.1% | – |
| GDP.pdf (AA) | Artificial Analysis | 22.8% | – |
| AA-Briefcase v1.1 | Artificial Analysis | 1520 | – |
| Terminal-Bench-Science 0.1 (AA) | Artificial Analysis | 20.0% | – |
| Maths | |||
| AIME 2026 | MathArena | – | 91.7% |
| HMMT February 2026 | MathArena | – | 81.1% |
| MathArena Apex | MathArena | – | 2.1% |
| Arena Maths | Arena (LMArena) | – | 1423 |
| Knowledge | |||
| AA-Omniscience v1.1 | Artificial Analysis | 7.4 | – |
| Excel Modeling Benchmark v1.1 | Vals AI | 72.2% | – |
| AA-Omniscience v1.1: accuracy | Artificial Analysis | 21.1% | – |
| AA-Omniscience v1.1: non-hallucination | Artificial Analysis | 82.6% | – |
| Global-MMLU-Lite (AA) | Artificial Analysis | – | 83.8%† |
| Instruction following | |||
| IFBench | Artificial Analysis | – | 46.9%† |
| Arena Instruction Following | Arena (LMArena) | – | 1400 |
| Human preference | |||
| Arena Text | Arena (LMArena) | – | 1409 |
| Arena Creative Writing | Arena (LMArena) | – | 1358 |
| Arena Multi-turn | Arena (LMArena) | – | 1414 |
| Arena Longer Queries | Arena (LMArena) | – | 1422 |
| Multimodal | |||
| Arena Vision | Arena (LMArena) | – | 1240 |
| MMMU-Pro | Artificial Analysis | – | 70.0%† |
| Long context | |||
| AA-LCR v1.1 | Artificial Analysis | – | 64.0%† |
| MLCR | Artificial Analysis | 18.3% | – |
| Composite | |||
| AA Intelligence Index v5.0.1 | Artificial Analysis | 36.0 | – |
| AA Intelligence Index v4.3.2 | Artificial Analysis | – | 19.4† |
| AA Progress Index v1 | Artificial Analysis | – | 250.2† |
| Vals Index | Vals AI | 60.5 | – |
| Vals Finance Index | Vals AI | 60.9% | – |
| Vals Legal Index | Vals AI | 75.7% | – |
| Safety & honesty | |||
| FORTRESS | Scale AI SEAL | 11.9%† | – |
| Metadata | |||
| AA Openness Index | Artificial Analysis | – | 38.9† |
Data as of 2026-11-16. Best configuration of each model; every score links to its source on the model pages.
What changed
- Qwen3.5-27B: BenchLeader Index 52.5 → 52.0.
- Muse Ember (max): 1 new result (Arena Agent).
- Muse Ember (max): 2 new results (MLCR, Arena WebDev).
Related comparisons
- Muse Ember vs Muse Spark Open 816B A42B
- Muse Ember vs Mistral Large 4
- Muse Ember vs Gemini 3.1 Flash Lite
- Muse Ember vs Inkling Small
- Qwen3.5-27B vs MiniMax M3.1 Flash
- Muse Ember vs Gemma 4 31B
Rankings: Best LLM overall · Best LLM for agents · Best cheap LLMs
Muse Ember vs Qwen3.5-27B: questions
- Is Muse Ember better than Qwen3.5-27B?
- Not enough shared benchmarks for a verdict. The BenchLeader Index combines every independent quality benchmark; the two are within the margin of error as of 16 Nov 2026, so check the category scores for your use.
- Muse Ember vs Qwen3.5-27B: which is cheaper, and will that change?
- Muse Ember is cheaper today: $0.200 against $0.825 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
- Is Muse Ember or Qwen3.5-27B better for coding?
- Muse Ember scores higher in coding (60 vs 44 on the category index, where 50 is average). On Arena WebDev, the top shared coding benchmark, it is 1609 for Muse Ember against 1357 for Qwen3.5-27B.
- Which is faster, Muse Ember or Qwen3.5-27B?
- Muse Ember streams 3×+ faster (437 vs 38 output tokens per second, Artificial Analysis 7-day median).
- Which is better for long documents?
- Muse Ember accepts more context: 1,048,576 against 262,144 tokens for Qwen3.5-27B.
More questions (3)
- Which has lower latency?
- Qwen3.5-27B answers first: 2.9 s against 5.0 s to the first answer token, as Artificial Analysis measures it. Reasoning effort drives most of this: a higher setting thinks longer before it answers.
- Do independent benchmarks agree on Muse Ember vs Qwen3.5-27B?
- Across the 1 benchmarks both have, Muse Ember leads 1, Qwen3.5-27B leads 0.
- When was Muse Ember released?
- Muse Ember was released on 9 Nov 2026.