Muse Spark Open vs Qwen3.5-122B-A10B
Full profiles: Muse Spark Open (Meta) and Qwen3.5-122B-A10B (Alibaba): every benchmark, price and rank.
Updated · Data from 10 sources
As of 16 Nov 2026, Muse Spark Open and Qwen3.5-122B-A10B are neck and neck on quality (53.5 vs 54.1 on the BenchLeader Index, inside the ±8.9 margin), and Qwen3.5-122B-A10B is 1.8× cheaper at $1.10 vs $2.00 per 1M tokens blended. Choose Muse Spark Open for multimodal (63 vs 57), where it also streams ~2× faster; choose Qwen3.5-122B-A10B for fast first answers (1.1 s vs 9.9 s) and coding (45 vs 36) and reasoning (59 vs 51).
Key facts (as of 16 Nov 2026)
- Muse Spark Open scores 53.5 on the BenchLeader Index and Qwen3.5-122B-A10B scores 54.1 (BenchLeader, 16 Nov 2026).
- Muse Spark Open costs $2.00 per 1M tokens blended (list price, 16 Nov 2026).
- Qwen3.5-122B-A10B costs $1.10 per 1M tokens blended (list price, 16 Nov 2026).
- Muse Spark Open streams ~2× faster than Qwen3.5-122B-A10B (Artificial Analysis, 7-day median, 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 | Qwen3.5-122B-A10B | 45 vs 36 category score | High |
| Reasoning | Qwen3.5-122B-A10B | 59 vs 51 category score | High |
| Images and other media | Muse Spark Open | 63 vs 57 category score | Medium |
| Chat quality (human preference) | – | Only Qwen3.5-122B-A10B has an Arena rating (59 category score) | Not comparable |
| Lowest cost | Qwen3.5-122B-A10B | $1.10 vs $2.00 per 1M blended; agentic coding session $0.037 vs $0.092 | – |
| Fastest first answer | Qwen3.5-122B-A10B | 1.1 s (Qwen3.5-122B-A10B) vs 9.9 s (Muse Spark Open (xhigh)) 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
- Muse Spark Open (xhigh) and Qwen3.5-122B-A10B are neck and neck on quality (53.5 vs 54.1, within the ±8.9 margin).
- Muse Spark Open (xhigh) is stronger in multimodal.
- Qwen3.5-122B-A10B is stronger in coding, reasoning.
- Qwen3.5-122B-A10B is 1.8× cheaper ($1.10 vs $2.00 per 1M blended).
- Muse Spark Open (xhigh) streams 2.0× faster (180 vs 92 tokens per second).
At a glance
| Spec | Muse Spark Open | Qwen3.5-122B-A10B |
|---|---|---|
| Developer | Meta | Alibaba |
| Released | 15 Oct 2026 | 24 Feb 2026 |
| Context window (tokens) | 1,048,576 | 262,144 |
| Compared at | Muse Spark Open (xhigh) | Qwen3.5-122B-A10B |
| Open weights | Yes | Yes |
| Price in / out / cached input per 1M | $1.25 / $4.25 / $0.150 | $0.400 / $3.20 / – |
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 Spark Open (xhigh) | Qwen3.5-122B-A10B | Leader |
|---|---|---|---|
| Reasoning | 51.3 | 58.9 | Qwen3.5-122B-A10B |
| Coding | 36.0 | 45.2 | Qwen3.5-122B-A10B |
| Maths | – | 60.1 | not comparable |
| Knowledge | 58.8 | – | not comparable |
| Human preference | – | 59.2 | not comparable |
| Multimodal | 63.3 | 57.4 | Muse Spark Open |
| Long context | 63.8 | – | 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.
Not drawn, measured for fewer than two of these models: Agents & tools, Maths, Knowledge, Instruction following, Long context, Human preference.
Where sources disagree
Across the 0 benchmarks both have, Muse Spark Open leads 0, Qwen3.5-122B-A10B leads 0.
| Overall index | Muse Spark Open (xhigh) | Qwen3.5-122B-A10B | Leader |
|---|---|---|---|
| BenchLeader Index | 53.5 | 54.1 | 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 Spark Open (xhigh) | Qwen3.5-122B-A10B |
|---|---|---|
| Blended price per 1M tokens | $2.00 | $1.10 |
| Input / output per 1M | $1.25 / $4.25 | $0.400 / $3.20 |
| Cached input per 1M | $0.150 | – |
| Price per task: chat reply | $0.0018 | $0.0011 |
| Price per task: summarize a 30-page report | $0.018 | $0.0067 |
| Price per task: code edit | $0.014 | $0.0072 |
| Price per task: agentic coding session | $0.092 | $0.037 |
| Price per task: structured extraction | $0.0034 | $0.0014 |
| Agentic coding session, input mostly cached | $0.043 | – |
| Cost per benchmark task (AA Index) | $2.02 | – |
| Output tokens per task (AA Index) | 66k | – |
| Of which reasoning | 32k | – |
| Time per task (AA Index) | 6 min | – |
| Output speed | 180 tok/s | 92 tok/s |
| Time to first answer | 9.9 s | 1.1 s |
| Time to a complete answer, last measured 13 Nov 2026 | 39.9 s | – |
Each model is compared at its best-ranked reasoning-effort setting. Muse Spark Open (xhigh) scores 53.5 and takes 10 s to its first answer token. Qwen3.5-122B-A10B scores 54.1 and takes 1 s to its first answer token; Qwen3.5-122B-A10B (thinking) scores 53.9 in 18 s, Qwen3.5-122B-A10B (no reasoning) scores 48.4 in 4 s. 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
93 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 Spark Open (xhigh) | Qwen3.5-122B-A10B |
|---|---|---|---|
| BenchLeader Index | BenchLeader | 53.5 | 54.1 |
| Coding score | BenchLeader | 36.0 | 45.2 |
| Knowledge score | BenchLeader | 58.8 | – |
| Long context score | BenchLeader | 63.8 | – |
| Multimodal score | BenchLeader | 63.3 | 57.4 |
| Reasoning score | BenchLeader | 51.3 | 58.9 |
| Human preference score | BenchLeader | – | 59.2 |
| Maths score | BenchLeader | – | 60.1 |
| Blended price $/M | BenchLeader | $2.00 | $1.10 |
| Output speed | BenchLeader | 180 tok/s | 92 tok/s |
| Time to first answer | BenchLeader | 9.9 s | 1.1 s |
| Context window | BenchLeader | 1,048,576 tokens | 262,144 tokens |
| Reasoning | |||
| Arena Hard Prompts | Arena (LMArena) | – | 1433 |
| LiveBench 2.0 Reasoning | LiveBench | 73.7%† | – |
| GPQA Diamond (AA) | Artificial Analysis | 89.0% | 85.7%† |
| Humanity's Last Exam (AA) | Artificial Analysis | 43.9% | 25.2%† |
| HLE-Diamond (AA) | Artificial Analysis | 26.3% | 7.3%† |
| ARC-AGI-1 | ARC Prize | 89.0% | – |
| ARC-AGI-2 | ARC Prize | 57.5% | – |
| ARC-AGI-3 | ARC Prize | 0.4% | – |
| CritPt | Artificial Analysis | 14.9% | 0.6%† |
| CritPt v2 hard subset (AA) | Artificial Analysis | 29.5% | – |
| HLE-Diamond (Scale) | Scale AI SEAL | 28.0%† | – |
| Arena Expert | Arena (LMArena) | – | 1441 |
| Mystery Game Puzzles | Epoch AI Benchmarking Hub | – | 17.0%† |
| LMCA | Epoch AI Benchmarking Hub | – | 32.2%† |
| DTBench | Epoch AI Benchmarking Hub | – | 84.3%† |
| Coding | |||
| SciCode | SciCode | – | 35.6%† |
| Arena Coding | Arena (LMArena) | – | 1458 |
| Arena WebDev | Arena (LMArena) | 1526† | 1360 |
| LiveBench 2.0 Coding | LiveBench | 55.1%† | – |
| SciCode (AA) | Artificial Analysis | 58.5% | 39.7%† |
| Code Migration | Vals AI | 21.2% | – |
| Code Migration v1.1 | Vals AI | 26.9% | – |
| Vibe Code Bench v1.1 | Vals AI | 76.2% | – |
| SWE Atlas v1.1 | Scale AI SEAL | 42.3%† | – |
| SWE Atlas v1.1: Codebase QnA | Scale AI SEAL | 50.0%† | – |
Show all 93 benchmarks
| Metric | Publisher | Muse Spark Open (xhigh) | Qwen3.5-122B-A10B |
|---|---|---|---|
| Coding | |||
| SWE Atlas v1.1: Test Writing | Scale AI SEAL | 41.1%† | – |
| SWE Atlas v1.1: Refactoring | Scale AI SEAL | 35.7%† | – |
| SWE-Bench Pro V2 | Scale AI SEAL | 95.0% | – |
| SWE-Bench Pro V2 HARD | Scale AI SEAL | 82.3% | – |
| Terminal-Bench 4.0 (AA) | Artificial Analysis | 8.1% | 0.0%† |
| Terminal-Bench 2.1 (AA) | Artificial Analysis | 76.4% | 47.6%† |
| GameBench 2 Overall | DuelLab GameBench 2 | 55.3† | – |
| GameBench 2 by reasoning setting | DuelLab GameBench 2 | 59.3 | – |
| Design Arena Website | Design Arena | 1305 | – |
| FrontierCode 1.1 (AA) | Artificial Analysis | 27.6% | – |
| FrontierCode 2.0 (AA) | Artificial Analysis | 7.6% | – |
| Agents & tools | |||
| LiveBench Long Agentic | LiveBench | 12.1%† | – |
| LiveBench Browser | LiveBench | 26.2%† | – |
| Terminal-Bench Hard | Artificial Analysis | – | 31.1%† |
| τ²-Bench Telecom (AA) | Artificial Analysis | – | 93.6%† |
| HiL-Bench | Scale AI SEAL | 59.7%† | – |
| GDPval-AA v2.1, normalised score | Artificial Analysis | 45.1% | 15.7%† |
| GDPval-AA v2.1 | Artificial Analysis | 1401 | 821† |
| τ³-Banking (AA) | Artificial Analysis | – | 15.3%† |
| Finance Agent v2 | Vals AI | 61.3% | – |
| Finance Agent v2.1 | Vals AI | 60.7% | – |
| Harvey's Legal Agent Benchmark v1.1 | Vals AI | 25.0% | – |
| Harvey's Legal Agent Benchmark v1 | Vals AI | 16.7% | – |
| Legal Research Bench | Vals AI | 38.5% | – |
| Tax Agent Bench | Vals AI | 26.3% | – |
| Terminal-Bench 4.0 (Vals) | Vals AI | 6.6% | – |
| AutomationBench | Artificial Analysis | 49.7% | 5.5%† |
| GDP.pdf (AA) | Artificial Analysis | 18.6% | 3.4%† |
| AA-Briefcase v1.1 | Artificial Analysis | 1280 | 499† |
| GDPx | GDPx by SemiAnalysis | 1.7%† | – |
| Terminal-Bench-Science 0.1 (AA) | Artificial Analysis | 1.9% | – |
| Maths | |||
| ArXivMath | MathArena | 27.0% | – |
| BrokenArXiv | MathArena | 30.1% | – |
| Arena Maths | Arena (LMArena) | – | 1425 |
| Knowledge | |||
| LiveBench 2.0 Data Analysis | LiveBench | 60.2%† | – |
| LiveBench Research | LiveBench | 82.7%† | – |
| AA-Omniscience v1.1 | Artificial Analysis | 25.8 | – |
| PRBench Legal | Scale AI SEAL | 59.8%† | – |
| Excel Modeling Benchmark | Vals AI | 53.7% | – |
| Excel Modeling Benchmark v1.1 | Vals AI | 62.0% | – |
| AA-Omniscience v1.1: accuracy | Artificial Analysis | 42.0% | – |
| AA-Omniscience v1.1: non-hallucination | Artificial Analysis | 72.0% | – |
| Global-MMLU-Lite (AA) | Artificial Analysis | – | 89.1%† |
| Instruction following | |||
| IFBench | Artificial Analysis | – | 75.7%† |
| MultiChallenge | Scale AI SEAL | 74.2%† | – |
| Arena Instruction Following | Arena (LMArena) | – | 1406 |
| Human preference | |||
| Arena Text | Arena (LMArena) | – | 1416 |
| Arena Creative Writing | Arena (LMArena) | – | 1365 |
| Arena Multi-turn | Arena (LMArena) | – | 1419 |
| Arena Longer Queries | Arena (LMArena) | – | 1418 |
| Multimodal | |||
| Arena Vision | Arena (LMArena) | – | 1244 |
| LiveBench Design | LiveBench | 58.1%† | – |
| MMMU-Pro | Artificial Analysis | 80.9% | 75.0%† |
| SAGE | Vals AI | 85.8% | – |
| Humanity's Sixth Sense | Scale AI SEAL | 29.6%† | – |
| Long context | |||
| AA-LCR v1.1 | Artificial Analysis | 78.3% | 76.3%† |
| MLCR | Artificial Analysis | 7.8% | 7.8%† |
| Composite | |||
| Epoch Capabilities Index | Epoch AI Benchmarking Hub | 155.4† | – |
| LiveBench 2.0 | LiveBench | 48.5%† | – |
| AA Intelligence Index v5.0.1 | Artificial Analysis | 24.9 | – |
| AA Intelligence Index v4.3.2 | Artificial Analysis | 38.9 | 15.6† |
| AA Progress Index v1 | Artificial Analysis | 306.7 | 261.5† |
| Vals Index | Vals AI | 48.6 | – |
| Vals Finance Index | Vals AI | 53.0% | – |
| Vals Legal Index | Vals AI | 68.5% | – |
| Safety & honesty | |||
| FORTRESS | Scale AI SEAL | 7.6%† | – |
| DistressBench | Scale AI SEAL | 87.7%† | – |
| Metadata | |||
| AA Openness Index | Artificial Analysis | 27.8 | 38.9† |
Data as of 2026-11-16. Best configuration of each model; every score links to its source on the model pages.
DuelLab GameBench 2, release gb2-20261111-7b598bb13de4 (https://benchmarks.duellab.org/releases/gb2-20261111-7b598bb13de4/), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Scores mapped to BenchLeader model names; not endorsed by DuelLab.
What changed
- Muse Spark Open (xhigh): 1 score revision.
- Muse Spark Open (xhigh): 1 new result (GameBench 2 by reasoning setting); 1 score revision.
- Muse Spark Open (xhigh): BenchLeader Index 53.3 → 53.8.
- Muse Spark Open (xhigh): 1 new result (ARC-AGI-3).
- Muse Spark Open (xhigh): BenchLeader Index 59.5 → 53.3.
- Muse Spark Open (xhigh): 6 new results (AA Intelligence Index v5.0.1, AA Progress Index v1, HLE-Diamond (AA)…); 3 score revisions; cost per task moved from $0.84/M to $2.02/M (139% up).
- Qwen3.5-122B-A10B: BenchLeader Index 54.6 → 54.1.
- Muse Spark Open (xhigh): 6 new results (MLCR, Code Migration v1.1, Excel Modeling Benchmark v1.1…); cached input price moved from $1.25/M to $0.15/M (88% down).
- Muse Spark Open (xhigh): 6 new results (FrontierCode 1.1 (AA), Terminal-Bench-Science 0.1 (AA), Vals Index…).
- Muse Spark Open (xhigh): 9 new results (Vals Index, Code Migration, Excel Modeling Benchmark…).
Related comparisons
- Muse Spark Open vs Claude Opus 5.5
- Muse Spark Open vs GPT-6 Astra
- Muse Spark Open vs GPT-6.1 Sol
- Muse Spark Open vs Claude Fable 5
- Qwen3.5-122B-A10B vs MiniMax M3.1 Flash
- Muse Spark Open vs Gemini 4 Argon
- Muse Spark Open vs GPT-5.6 Sol
- Muse Spark Open vs Claude Sonnet 5.5
Rankings: Best LLM overall · Best LLM for agents · Best cheap LLMs
Muse Spark Open vs Qwen3.5-122B-A10B: questions
- Is Muse Spark Open better than Qwen3.5-122B-A10B?
- Muse Spark Open (xhigh) and Qwen3.5-122B-A10B are neck and neck on quality (53.5 vs 54.1, within the ±8.9 margin). 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 Spark Open vs Qwen3.5-122B-A10B: which is cheaper, and will that change?
- Qwen3.5-122B-A10B is cheaper today: $1.10 against $2.00 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
- Is Muse Spark Open or Qwen3.5-122B-A10B better for coding?
- Qwen3.5-122B-A10B scores higher in coding (45 vs 36 on the category index, where 50 is average).
- Which is faster, Muse Spark Open or Qwen3.5-122B-A10B?
- Muse Spark Open streams ~2× faster (180 vs 92 output tokens per second, Artificial Analysis 7-day median).
- Which is better for long documents?
- Muse Spark Open accepts more context: 1,048,576 against 262,144 tokens for Qwen3.5-122B-A10B.
More questions (2)
- Which has lower latency?
- Qwen3.5-122B-A10B answers first: 1.1 s against 9.9 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.
- When was Muse Spark Open released?
- Muse Spark Open was released on 15 Oct 2026.