Model comparison
MiniMax-M2.5 vs Muse Spark 1.3
Muse Spark 1.3 is the stronger model overall, scoring 54.4 to 38.1 on the Noometry Index. MiniMax-M2.5 costs 3.8× less per token, which makes it the better buy when Muse Spark 1.3's lead doesn't matter for your workload.
Last verified . 19 shared benchmarks.
MiniMax
MetaSummary
- They share 19 benchmarks with published results for both. MiniMax-M2.5 scores higher in 0 categories and Muse Spark 1.3 in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Muse Spark 1.3 leads 72.4 to 25.8.
- The biggest single-benchmark swing is NYT Connections (extended): 16.8% for MiniMax-M2.5 and 85.1% for Muse Spark 1.3.
- MiniMax-M2.5 is cheaper at $0.30 / $1.20 per million input/output tokens, against $1.25 / $4.25 for Muse Spark 1.3.
- Muse Spark 1.3 accepts more context: 1.05M tokens versus 205K.
- MiniMax-M2.5 has downloadable open weights; the other is API-only.
Side by side
| MiniMax-M2.5 | Muse Spark 1.3 | |
|---|---|---|
| Provider | MiniMax | Meta |
| Noometry Index | 38.1 | 54.4 |
| Released | 2026-02-12 | 2026-09-02 |
| Weights | Open | Proprietary |
| Context window | 205K | 1.05M |
| Max output | 131K | 131K |
| Input $ / M tokens | $0.30 | $1.25 |
| Output $ / M tokens | $1.20 | $4.25 |
| Results tracked | 33 | 34 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Muse Spark 1.3 leads
MiniMax-M2.5: 47.9 (#53), Muse Spark 1.3: 56.1 (#18)
| Benchmark | MiniMax-M2.5 | Muse Spark 1.3 |
|---|---|---|
| LMArena WebDev | 1386 | 1655 |
| LMArena Coding | 1380 | 1516 |
| SWE-bench Verified (bash only) | 75.8% | — |
| CursorBench | — | 41.6% |
| SWE-bench Multilingual | 68.3% | — |
| SciCode | — | 59.7% |
| ALE-Bench | 618.17 | — |
Agentic & Tool Use Muse Spark 1.3 leads
MiniMax-M2.5: 30.2 (#70), Muse Spark 1.3: 38.6 (#28)
| Benchmark | MiniMax-M2.5 | Muse Spark 1.3 |
|---|---|---|
| Terminal-Bench | 42.7% | — |
| APEX-Agents | — | 57.8% |
| GDP.pdf | — | 27.6% |
| Vending-Bench 2 | -23.16 | — |
Reasoning Muse Spark 1.3 leads
MiniMax-M2.5: 17.6 (#281), Muse Spark 1.3: 52.9 (#27)
| Benchmark | MiniMax-M2.5 | Muse Spark 1.3 |
|---|---|---|
| NYT Connections (extended) | 16.8% | 85.1% |
| LMArena Hard Prompts | 1372 | 1503 |
| Epoch Capabilities Index | 146.68 | 156.89 |
| ARC-AGI-2 | 4.9% | — |
| Kagi LLM Benchmark | 55.2% | — |
| ARC-AGI-1 | 63.7% | — |
| CritPt | — | 24.9% |
| Chess Puzzles | — | 38% |
| Mystery Game Puzzles | — | 25% |
| DTBench | — | 96.5% |
| LMCA | — | 53.9% |
| Bench to the Future 3 | — | 0.14 |
Math Muse Spark 1.3 leads
MiniMax-M2.5: 25.8 (#242), Muse Spark 1.3: 72.4 (#16)
| Benchmark | MiniMax-M2.5 | Muse Spark 1.3 |
|---|---|---|
| LMArena Math | 1378 | 1497 |
| FrontierMath (Tiers 1-3) | — | 74.4% |
| FrontierMath Tier 4 | — | 46.3% |
| OTIS Mock AIME 2024-2025 | — | 99.2% |
| ProofBench | 4% | — |
Knowledge Muse Spark 1.3 leads
MiniMax-M2.5: 38.6 (#132), Muse Spark 1.3: 42.0 (#92)
| Benchmark | MiniMax-M2.5 | Muse Spark 1.3 |
|---|---|---|
| LMArena Expert | 1378 | 1516 |
| Vectara Hallucination Rate | 9.1% | — |
Multimodal Not comparable
MiniMax-M2.5: —, Muse Spark 1.3: 43.0 (#21)
| Benchmark | MiniMax-M2.5 | Muse Spark 1.3 |
|---|---|---|
| LMArena Vision | — | 1310 |
| LMArena Document | — | 1471 |
Multilingual Muse Spark 1.3 leads
MiniMax-M2.5: 47.3 (#147), Muse Spark 1.3: 57.8 (#6)
| Benchmark | MiniMax-M2.5 | Muse Spark 1.3 |
|---|---|---|
| LMArena Non-English | 1338 | 1485 |
| LMArena Chinese | 1392 | 1527 |
| LMArena French | 1363 | 1534 |
| LMArena Korean | 1232 | 1513 |
| LMArena Russian | 1359 | 1491 |
| LMArena Spanish | 1354 | 1485 |
| LMArena German | 1361 | — |
| LMArena Japanese | 1169 | — |
Instruction Following Muse Spark 1.3 leads
MiniMax-M2.5: 71.7 (#143), Muse Spark 1.3: 77.8 (#14)
| Benchmark | MiniMax-M2.5 | Muse Spark 1.3 |
|---|---|---|
| LMArena Instruction Following | 1353 | 1481 |
Long Context Muse Spark 1.3 leads
MiniMax-M2.5: 37.4 (#211), Muse Spark 1.3: 45.7 (#29)
| Benchmark | MiniMax-M2.5 | Muse Spark 1.3 |
|---|---|---|
| LMArena Longer Query | 1367 | 1491 |
| CL-bench | 11.4% | — |
| CL-bench Life | 6.3% | — |
Writing & Preference Muse Spark 1.3 leads
MiniMax-M2.5: 54.1 (#149), Muse Spark 1.3: 73.7 (#10)
| Benchmark | MiniMax-M2.5 | Muse Spark 1.3 |
|---|---|---|
| LMArena Text | 1359 | 1490 |
| LMArena Creative Writing | 1331 | 1450 |
| EQ-Bench Creative Writing | 1361 | 1906 |
| LMArena Multi-Turn | 1364 | 1486 |
Frequently asked questions
Is MiniMax-M2.5 better than Muse Spark 1.3?
Muse Spark 1.3 is the stronger model overall, scoring 54.4 to 38.1 on the Noometry Index. MiniMax-M2.5 costs 3.8× less per token, which makes it the better buy when Muse Spark 1.3's lead doesn't matter for your workload.
Which is cheaper, MiniMax-M2.5 or Muse Spark 1.3?
MiniMax-M2.5 is cheaper. It lists at $0.30 per million input tokens and $1.20 per million output tokens; Muse Spark 1.3 lists at $1.25 and $4.25.
Is MiniMax-M2.5 or Muse Spark 1.3 better for coding?
Muse Spark 1.3 scores higher on coding benchmarks: 56.1 versus 47.9 in the Noometry coding category.
Which has the bigger context window?
Muse Spark 1.3 does, with 1.05M tokens against 205K.
How many benchmarks do MiniMax-M2.5 and Muse Spark 1.3 share?
19 benchmarks have published results for both models. MiniMax-M2.5 has 33 scored results on Noometry and Muse Spark 1.3 has 34.