Model comparison
Grok-3 mini vs Muse Spark
Muse Spark is the stronger model overall, scoring 50.6 to 41.2 on the Noometry Index.
Last verified . 20 shared benchmarks.
MetaSummary
- They share 20 benchmarks with published results for both. Grok-3 mini scores higher in 1 category and Muse Spark in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Muse Spark leads 35.9 to 13.6.
- The biggest single-benchmark swing is GPQA Diamond: 76.3% for Grok-3 mini and 89.8% for Muse Spark.
Side by side
| Grok-3 mini | Muse Spark | |
|---|---|---|
| Provider | xAI | Meta |
| Noometry Index | 41.2 | 50.6 |
| Released | 2025-04-09 | 2026-04-08 |
| Weights | Proprietary | Proprietary |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 35 | 27 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Muse Spark leads
Grok-3 mini: 40.8 (#131), Muse Spark: 46.2 (#69)
| Benchmark | Grok-3 mini | Muse Spark |
|---|---|---|
| LMArena Coding | 1379 | 1481 |
| Aider Polyglot | 49.3% | — |
| SciCode | — | 51.5% |
| WeirdML | 42.6% | — |
Reasoning Muse Spark leads
Grok-3 mini: 13.6 (#334), Muse Spark: 35.9 (#67)
| Benchmark | Grok-3 mini | Muse Spark |
|---|---|---|
| LMArena Hard Prompts | 1375 | 1474 |
| Epoch Capabilities Index | 140.35 | 152.04 |
| ARC-AGI-2 | 0.4% | — |
| Kagi LLM Benchmark | 61.3% | — |
| ARC-AGI-1 | 16.5% | — |
| CritPt | — | 11.3% |
Math Muse Spark leads
Grok-3 mini: 42.1 (#85), Muse Spark: 47.8 (#66)
| Benchmark | Grok-3 mini | Muse Spark |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 77.8% | 88.9% |
| LMArena Math | 1386 | 1455 |
| FrontierMath (Feb 2025 set) | 5.9% | 39% |
| ProofBench | — | 17% |
| Omni-MATH | 31.8% | — |
| MATH Level 5 | 90.9% | — |
| FrontierMath Tier 4 (v1) | — | 14.6% |
Knowledge Muse Spark leads
Grok-3 mini: 46.4 (#81), Muse Spark: 65.7 (#13)
| Benchmark | Grok-3 mini | Muse Spark |
|---|---|---|
| GPQA Diamond | 76.3% | 89.8% |
| LMArena Expert | 1395 | 1457 |
| Humanity's Last Exam | — | 40.6% |
| MMLU-Pro | 79.9% | — |
| Confabulations | 10.8% | — |
| GPQA (HELM) | 67.5% | — |
Multimodal Not comparable
Grok-3 mini: —, Muse Spark: 43.4 (#24)
| Benchmark | Grok-3 mini | Muse Spark |
|---|---|---|
| LMArena Vision | — | 1306 |
| LMArena Document | — | 1444 |
Multilingual Muse Spark leads
Grok-3 mini: 48.1 (#145), Muse Spark: 56.1 (#24)
| Benchmark | Grok-3 mini | Muse Spark |
|---|---|---|
| LMArena Non-English | 1352 | 1464 |
| LMArena Chinese | 1387 | 1509 |
| LMArena French | 1357 | 1497 |
| LMArena German | 1349 | 1497 |
| LMArena Korean | 1335 | 1459 |
| LMArena Russian | 1353 | 1466 |
| LMArena Spanish | 1381 | 1472 |
| LMArena Japanese | 1342 | — |
Instruction Following Grok-3 mini leads
Grok-3 mini: 78.5 (#9), Muse Spark: 75.9 (#51)
| Benchmark | Grok-3 mini | Muse Spark |
|---|---|---|
| LMArena Instruction Following | 1357 | 1442 |
| IFEval | 95.1% | — |
Long Context Muse Spark leads
Grok-3 mini: 41.0 (#147), Muse Spark: 44.4 (#69)
| Benchmark | Grok-3 mini | Muse Spark |
|---|---|---|
| LMArena Longer Query | 1372 | 1451 |
| Fiction.LiveBench | 66.7% | — |
Writing & Preference Muse Spark leads
Grok-3 mini: 52.5 (#169), Muse Spark: 66.0 (#39)
| Benchmark | Grok-3 mini | Muse Spark |
|---|---|---|
| LMArena Text | 1370 | 1474 |
| LMArena Creative Writing | 1342 | 1459 |
| LMArena Multi-Turn | 1355 | 1477 |
| Short-Story Creative Writing | 73.5% | — |
| WildBench | 65.1% | — |
Frequently asked questions
Is Grok-3 mini better than Muse Spark?
Muse Spark is the stronger model overall, scoring 50.6 to 41.2 on the Noometry Index.
Is Grok-3 mini or Muse Spark better for coding?
Muse Spark scores higher on coding benchmarks: 46.2 versus 40.8 in the Noometry coding category.
How many benchmarks do Grok-3 mini and Muse Spark share?
20 benchmarks have published results for both models. Grok-3 mini has 35 scored results on Noometry and Muse Spark has 27.