Model comparison
Muse Spark 1.2 vs Qwen2.5 7B Instruct
Muse Spark 1.2 is the stronger model overall, scoring 49.8 to 28.6 on the Noometry Index. Qwen2.5 7B Instruct costs 6.5× less per token, which makes it the better buy when Muse Spark 1.2's lead doesn't matter for your workload.
Last verified . 3 shared benchmarks.
Meta
Alibaba (Qwen)Summary
- They share 3 benchmarks with published results for both. Muse Spark 1.2 scores higher in 7 categories and Qwen2.5 7B Instruct in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Muse Spark 1.2 leads 53.9 to 17.0.
- The biggest single-benchmark swing is DTBench: 94.7% for Muse Spark 1.2 and 47.7% for Qwen2.5 7B Instruct.
- Qwen2.5 7B Instruct is cheaper at $0.17 / $0.70 per million input/output tokens, against $1.25 / $4.25 for Muse Spark 1.2.
- Muse Spark 1.2 accepts more context: 1.05M tokens versus 131K.
- Qwen2.5 7B Instruct has downloadable open weights; the other is API-only.
Side by side
| Muse Spark 1.2 | Qwen2.5 7B Instruct | |
|---|---|---|
| Provider | Meta | Alibaba (Qwen) |
| Noometry Index | 49.8 | 28.6 |
| Released | 2026-08-05 | 2024-09 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 131K |
| Max output | 131K | 8K |
| Input $ / M tokens | $1.25 | $0.17 |
| Output $ / M tokens | $4.25 | $0.70 |
| Results tracked | 28 | 15 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Muse Spark 1.2 leads
Muse Spark 1.2: 48.6 (#49), Qwen2.5 7B Instruct: 35.9 (#204)
| Benchmark | Muse Spark 1.2 | Qwen2.5 7B Instruct |
|---|---|---|
| DeepSWE | 54.9% | — |
| LMArena WebDev | 1532 | — |
| FrontierSWE | 12% | — |
| SciCode | 56.4% | — |
| WeirdML | 60.3% | — |
| BigCodeBench Instruct | — | 37.6% |
| LMArena Coding | 1497 | — |
| BigCodeBench Complete | — | 46.1% |
Agentic & Tool Use Muse Spark 1.2 leads
Muse Spark 1.2: 29.5 (#77), Qwen2.5 7B Instruct: 23.5 (#117)
| Benchmark | Muse Spark 1.2 | Qwen2.5 7B Instruct |
|---|---|---|
| APEX-Agents | 36.4% | — |
| BALROG | — | 7.8% |
| GDP.pdf | 16% | — |
Reasoning Muse Spark 1.2 leads
Muse Spark 1.2: 50.4 (#32), Qwen2.5 7B Instruct: 14.3 (#317)
| Benchmark | Muse Spark 1.2 | Qwen2.5 7B Instruct |
|---|---|---|
| DTBench | 94.7% | 47.7% |
| LMCA | 48.4% | 6.4% |
| Epoch Capabilities Index | 155.18 | 118.44 |
| SimpleBench | 74.5% | — |
| NYT Connections (extended) | 79.2% | — |
| CritPt | 17.7% | — |
| Chess Puzzles | — | 0% |
| LMArena Hard Prompts | 1490 | — |
Math Muse Spark 1.2 leads
Muse Spark 1.2: 43.8 (#68), Qwen2.5 7B Instruct: 11.8 (#300)
| Benchmark | Muse Spark 1.2 | Qwen2.5 7B Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | — | 2.5% |
| ProofBench | 43% | — |
| Omni-MATH | — | 29.4% |
Knowledge Muse Spark 1.2 leads
Muse Spark 1.2: 53.9 (#49), Qwen2.5 7B Instruct: 17.0 (#281)
| Benchmark | Muse Spark 1.2 | Qwen2.5 7B Instruct |
|---|---|---|
| GPQA Diamond | — | 35.5% |
| SimpleQA Verified | 60.3% | — |
| MMLU-Pro | — | 53.9% |
| GPQA (HELM) | — | 34.1% |
| LMArena Expert | 1485 | — |
| MMLU | — | 72.9% |
Multimodal Not comparable
Muse Spark 1.2: 42.6 (#24), Qwen2.5 7B Instruct: —
| Benchmark | Muse Spark 1.2 | Qwen2.5 7B Instruct |
|---|---|---|
| LMArena Vision | 1304 | — |
Multilingual Not comparable
Muse Spark 1.2: 57.4 (#10), Qwen2.5 7B Instruct: —
| Benchmark | Muse Spark 1.2 | Qwen2.5 7B Instruct |
|---|---|---|
| LMArena Non-English | 1479 | — |
| LMArena Chinese | 1519 | — |
| LMArena Russian | 1494 | — |
Instruction Following Muse Spark 1.2 leads
Muse Spark 1.2: 77.1 (#30), Qwen2.5 7B Instruct: 63.4 (#226)
| Benchmark | Muse Spark 1.2 | Qwen2.5 7B Instruct |
|---|---|---|
| IFEval | — | 74.1% |
| LMArena Instruction Following | 1466 | — |
Long Context Not comparable
Muse Spark 1.2: 45.4 (#35), Qwen2.5 7B Instruct: —
| Benchmark | Muse Spark 1.2 | Qwen2.5 7B Instruct |
|---|---|---|
| LMArena Longer Query | 1483 | — |
Writing & Preference Muse Spark 1.2 leads
Muse Spark 1.2: 72.6 (#14), Qwen2.5 7B Instruct: 48.8 (#194)
| Benchmark | Muse Spark 1.2 | Qwen2.5 7B Instruct |
|---|---|---|
| LMArena Text | 1486 | — |
| LMArena Creative Writing | 1444 | — |
| EQ-Bench Creative Writing | 1840 | — |
| WildBench | — | 73.1% |
| LMArena Multi-Turn | 1504 | — |
Frequently asked questions
Is Muse Spark 1.2 better than Qwen2.5 7B Instruct?
Muse Spark 1.2 is the stronger model overall, scoring 49.8 to 28.6 on the Noometry Index. Qwen2.5 7B Instruct costs 6.5× less per token, which makes it the better buy when Muse Spark 1.2's lead doesn't matter for your workload.
Which is cheaper, Muse Spark 1.2 or Qwen2.5 7B Instruct?
Qwen2.5 7B Instruct is cheaper. It lists at $0.17 per million input tokens and $0.70 per million output tokens; Muse Spark 1.2 lists at $1.25 and $4.25.
Is Muse Spark 1.2 or Qwen2.5 7B Instruct better for coding?
Muse Spark 1.2 scores higher on coding benchmarks: 48.6 versus 35.9 in the Noometry coding category.
Which has the bigger context window?
Muse Spark 1.2 does, with 1.05M tokens against 131K.
How many benchmarks do Muse Spark 1.2 and Qwen2.5 7B Instruct share?
3 benchmarks have published results for both models. Muse Spark 1.2 has 28 scored results on Noometry and Qwen2.5 7B Instruct has 15.