Model comparison
Muse Spark 1.3 vs Qwen3.8 Max
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 54.8 on the Noometry Index. Muse Spark 1.3 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 . 33 shared benchmarks.
Meta
Alibaba (Qwen)Summary
- They share 33 benchmarks with published results for both. Muse Spark 1.3 scores higher in 4 categories and Qwen3.8 Max in 6 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3.8 Max leads 61.7 to 42.6.
- The biggest single-benchmark swing is Mystery Game Puzzles: 25% for Muse Spark 1.3 and 38% for Qwen3.8 Max.
- Muse Spark 1.3 is cheaper at $1.25 / $4.25 per million input/output tokens, against $2 / $6 for Qwen3.8 Max.
- Muse Spark 1.3 accepts more context: 1.05M tokens versus 1M.
Side by side
| Muse Spark 1.3 | Qwen3.8 Max | |
|---|---|---|
| Provider | Meta | Alibaba (Qwen) |
| Noometry Index | 54.8 | 56.8 |
| Released | 2026-09-02 | 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 Muse Spark 1.3 leads
Muse Spark 1.3: 56.6 (#21), Qwen3.8 Max: 53.5 (#29)
| Benchmark | Muse Spark 1.3 | Qwen3.8 Max |
|---|---|---|
| LMArena WebDev | 1657 | 1674 |
| SciCode | 59.7% | 53.2% |
| LMArena Coding | 1514 | 1502 |
| DeepSWE | — | 57.5% |
| CursorBench | 41.6% | — |
| FrontierSWE | — | 17.8% |
Agentic & Tool Use Qwen3.8 Max leads
Muse Spark 1.3: 38.6 (#30), Qwen3.8 Max: 45.4 (#14)
| Benchmark | Muse Spark 1.3 | Qwen3.8 Max |
|---|---|---|
| APEX-Agents | 57.8% | 63.3% |
| GDP.pdf | 27.6% | 23.2% |
| τ²-bench Banking | — | 55.1% |
Reasoning Too close to call
Muse Spark 1.3: 54.0 (#27), Qwen3.8 Max: 54.4 (#26)
| Benchmark | Muse Spark 1.3 | Qwen3.8 Max |
|---|---|---|
| NYT Connections (extended) | 85.1% | 88.3% |
| CritPt | 26% | 20% |
| Chess Puzzles | 38% | 40% |
| LMArena Hard Prompts | 1503 | 1496 |
| Mystery Game Puzzles | 25% | 38% |
| DTBench | 96.5% | 92% |
| LMCA | 53.9% | 46.2% |
| Epoch Capabilities Index | 156.75 | 156.41 |
| Bench to the Future 3 | 0.14 | — |
Math Too close to call
Muse Spark 1.3: 73.1 (#21), Qwen3.8 Max: 73.2 (#20)
| Benchmark | Muse Spark 1.3 | Qwen3.8 Max |
|---|---|---|
| FrontierMath (Tiers 1-3) | 74.4% | 74.7% |
| FrontierMath Tier 4 | 46.3% | 46.3% |
| OTIS Mock AIME 2024-2025 | 99.2% | 100% |
| ProofBench | 58% | 58% |
| LMArena Math | 1494 | 1499 |
Knowledge Qwen3.8 Max leads
Muse Spark 1.3: 42.6 (#95), Qwen3.8 Max: 61.7 (#27)
| Benchmark | Muse Spark 1.3 | Qwen3.8 Max |
|---|---|---|
| LMArena Expert | 1516 | 1507 |
| GPQA Diamond | — | 92.7% |
| SimpleQA Verified | — | 47.3% |
Multimodal Muse Spark 1.3 leads
Muse Spark 1.3: 43.7 (#22), Qwen3.8 Max: 37.2 (#75)
| Benchmark | Muse Spark 1.3 | Qwen3.8 Max |
|---|---|---|
| LMArena Vision | 1309 | 1314 |
| Furniture Assembly | — | 20% |
| LMArena Document | 1471 | — |
Multilingual Too close to call
Muse Spark 1.3: 57.4 (#8), Qwen3.8 Max: 56.7 (#18)
| Benchmark | Muse Spark 1.3 | Qwen3.8 Max |
|---|---|---|
| LMArena Non-English | 1481 | 1472 |
| LMArena Chinese | 1529 | 1538 |
| LMArena French | 1524 | 1503 |
| LMArena German | 1515 | 1483 |
| LMArena Japanese | 1474 | 1467 |
| LMArena Korean | 1501 | 1461 |
| LMArena Russian | 1490 | 1481 |
| LMArena Spanish | 1490 | 1492 |
Instruction Following Too close to call
Muse Spark 1.3: 77.5 (#22), Qwen3.8 Max: 77.6 (#17)
| Benchmark | Muse Spark 1.3 | Qwen3.8 Max |
|---|---|---|
| LMArena Instruction Following | 1477 | 1479 |
Long Context Too close to call
Muse Spark 1.3: 45.6 (#32), Qwen3.8 Max: 45.6 (#31)
| Benchmark | Muse Spark 1.3 | Qwen3.8 Max |
|---|---|---|
| LMArena Longer Query | 1488 | 1489 |
Writing & Preference Muse Spark 1.3 leads
Muse Spark 1.3: 73.6 (#9), Qwen3.8 Max: 67.1 (#30)
| Benchmark | Muse Spark 1.3 | Qwen3.8 Max |
|---|---|---|
| LMArena Text | 1490 | 1483 |
| LMArena Creative Writing | 1455 | 1479 |
| LMArena Multi-Turn | 1482 | 1489 |
| EQ-Bench Creative Writing | 1906 | — |
Frequently asked questions
Is Muse Spark 1.3 better than Qwen3.8 Max?
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 54.8 on the Noometry Index. Muse Spark 1.3 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 1.3 or Qwen3.8 Max?
Muse Spark 1.3 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 1.3 or Qwen3.8 Max better for coding?
Muse Spark 1.3 scores higher on coding benchmarks: 56.6 versus 53.5 in the Noometry coding category.
Which has the bigger context window?
Muse Spark 1.3 does, with 1.05M tokens against 1M.
How many benchmarks do Muse Spark 1.3 and Qwen3.8 Max share?
33 benchmarks have published results for both models. Muse Spark 1.3 has 37 scored results on Noometry and Qwen3.8 Max has 39.