Model comparison
Muse Spark 1.1 vs Qwen3.8 Max
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 49.9 on the Noometry Index. Muse Spark 1.1 costs 1.5× less per token, which makes it the better buy when Qwen3.8 Max's lead doesn't matter for your workload.
Last verified . 31 shared benchmarks.
Meta
Alibaba (Qwen)Summary
- They share 31 benchmarks with published results for both. Muse Spark 1.1 scores higher in 3 categories and Qwen3.8 Max in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.8 Max leads 73.2 to 45.5.
- The biggest single-benchmark swing is APEX-Agents: 31.8% for Muse Spark 1.1 and 63.3% for Qwen3.8 Max.
- Muse Spark 1.1 is cheaper at $1.25 / $4.25 per million input/output tokens, against $2 / $6 for Qwen3.8 Max.
- Muse Spark 1.1 accepts more context: 1.05M tokens versus 1M.
Side by side
| Muse Spark 1.1 | Qwen3.8 Max | |
|---|---|---|
| Provider | Meta | Alibaba (Qwen) |
| Noometry Index | 49.9 | 56.8 |
| Released | 2026-04-08 | 2026-08-02 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 1M |
| Max output | 131K | 131K |
| Input $ / M tokens | $1.25 | $2 |
| Output $ / M tokens | $4.25 | $6 |
| Results tracked | 37 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 Max leads
Muse Spark 1.1: 51.3 (#40), Qwen3.8 Max: 53.5 (#29)
| Benchmark | Muse Spark 1.1 | Qwen3.8 Max |
|---|---|---|
| DeepSWE | 53.3% | 57.5% |
| LMArena WebDev | 1542 | 1674 |
| SciCode | 58.8% | 53.2% |
| LMArena Coding | 1498 | 1502 |
| FrontierSWE | — | 17.8% |
Agentic & Tool Use Qwen3.8 Max leads
Muse Spark 1.1: 30.8 (#73), Qwen3.8 Max: 45.4 (#14)
| Benchmark | Muse Spark 1.1 | Qwen3.8 Max |
|---|---|---|
| APEX-Agents | 31.8% | 63.3% |
| τ²-bench Banking | 40.5% | 55.1% |
| GDP.pdf | 15% | 23.2% |
| GBAEval | 7.9% | — |
| Vending-Bench 2 | 6,520 | — |
Reasoning Qwen3.8 Max leads
Muse Spark 1.1: 47.1 (#44), Qwen3.8 Max: 54.4 (#26)
| Benchmark | Muse Spark 1.1 | Qwen3.8 Max |
|---|---|---|
| NYT Connections (extended) | 84.9% | 88.3% |
| CritPt | 15.1% | 20% |
| LMArena Hard Prompts | 1486 | 1496 |
| DTBench | 94.4% | 92% |
| LMCA | 49.9% | 46.2% |
| Epoch Capabilities Index | 154.21 | 156.41 |
| Chess Puzzles | — | 40% |
| Mystery Game Puzzles | — | 38% |
| Surface Evolver Bench | 52.5% | — |
Math Qwen3.8 Max leads
Muse Spark 1.1: 45.5 (#76), Qwen3.8 Max: 73.2 (#20)
| Benchmark | Muse Spark 1.1 | Qwen3.8 Max |
|---|---|---|
| ProofBench | 39% | 58% |
| LMArena Math | 1483 | 1499 |
| FrontierMath (Tiers 1-3) | — | 74.7% |
| FrontierMath Tier 4 | — | 46.3% |
| OTIS Mock AIME 2024-2025 | — | 100% |
Knowledge Qwen3.8 Max leads
Muse Spark 1.1: 53.1 (#59), Qwen3.8 Max: 61.7 (#27)
| Benchmark | Muse Spark 1.1 | Qwen3.8 Max |
|---|---|---|
| SimpleQA Verified | 57.8% | 47.3% |
| LMArena Expert | 1478 | 1507 |
| GPQA Diamond | — | 92.7% |
Multimodal Muse Spark 1.1 leads
Muse Spark 1.1: 42.6 (#29), Qwen3.8 Max: 37.2 (#75)
| Benchmark | Muse Spark 1.1 | Qwen3.8 Max |
|---|---|---|
| LMArena Vision | 1293 | 1314 |
| Furniture Assembly | — | 20% |
| LMArena Document | 1465 | — |
Multilingual Too close to call
Muse Spark 1.1: 56.7 (#17), Qwen3.8 Max: 56.7 (#18)
| Benchmark | Muse Spark 1.1 | Qwen3.8 Max |
|---|---|---|
| LMArena Non-English | 1472 | 1472 |
| LMArena Chinese | 1518 | 1538 |
| LMArena French | 1494 | 1503 |
| LMArena German | 1466 | 1483 |
| LMArena Japanese | 1451 | 1467 |
| LMArena Korean | 1458 | 1461 |
| LMArena Russian | 1483 | 1481 |
| LMArena Spanish | 1464 | 1492 |
Instruction Following Qwen3.8 Max leads
Muse Spark 1.1: 76.5 (#39), Qwen3.8 Max: 77.6 (#17)
| Benchmark | Muse Spark 1.1 | Qwen3.8 Max |
|---|---|---|
| LMArena Instruction Following | 1457 | 1479 |
Long Context Too close to call
Muse Spark 1.1: 44.8 (#58), Qwen3.8 Max: 45.6 (#31)
| Benchmark | Muse Spark 1.1 | Qwen3.8 Max |
|---|---|---|
| LMArena Longer Query | 1462 | 1489 |
Writing & Preference Muse Spark 1.1 leads
Muse Spark 1.1: 73.4 (#11), Qwen3.8 Max: 67.1 (#30)
| Benchmark | Muse Spark 1.1 | Qwen3.8 Max |
|---|---|---|
| LMArena Text | 1479 | 1483 |
| LMArena Creative Writing | 1437 | 1479 |
| LMArena Multi-Turn | 1485 | 1489 |
| EQ-Bench Creative Writing | 1927 | — |
| EQ-Bench 4 | 1260 | — |
Frequently asked questions
Is Muse Spark 1.1 better than Qwen3.8 Max?
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 49.9 on the Noometry Index. Muse Spark 1.1 costs 1.5× less per token, which makes it the better buy when Qwen3.8 Max's lead doesn't matter for your workload.
Which is cheaper, Muse Spark 1.1 or Qwen3.8 Max?
Muse Spark 1.1 is cheaper. It lists at $1.25 per million input tokens and $4.25 per million output tokens; Qwen3.8 Max lists at $2 and $6.
Is Muse Spark 1.1 or Qwen3.8 Max better for coding?
Qwen3.8 Max scores higher on coding benchmarks: 53.5 versus 51.3 in the Noometry coding category.
Which has the bigger context window?
Muse Spark 1.1 does, with 1.05M tokens against 1M.
How many benchmarks do Muse Spark 1.1 and Qwen3.8 Max share?
31 benchmarks have published results for both models. Muse Spark 1.1 has 37 scored results on Noometry and Qwen3.8 Max has 39.