Model comparison
DeepSeek V4 Pro vs Muse Spark
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 50.6 on the Noometry Index.
Last verified . 22 shared benchmarks.
DeepSeek
MetaSummary
- They share 22 benchmarks with published results for both. DeepSeek V4 Pro scores higher in 5 categories and Muse Spark in 3 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek V4 Pro leads 56.5 to 35.9.
- The biggest single-benchmark swing is ProofBench: 50% for DeepSeek V4 Pro and 17% for Muse Spark.
- DeepSeek V4 Pro has downloadable open weights; the other is API-only.
Side by side
| DeepSeek V4 Pro | Muse Spark | |
|---|---|---|
| Provider | DeepSeek | Meta |
| Noometry Index | 54.3 | 50.6 |
| Released | 2026-04-24 | 2026-04-08 |
| Weights | Open | Proprietary |
| Context window | 1M | — |
| Max output | 393K | — |
| Input $ / M tokens | $0.66 | — |
| Output $ / M tokens | $1.98 | — |
| Results tracked | 48 | 27 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek V4 Pro leads
DeepSeek V4 Pro: 52.4 (#34), Muse Spark: 46.2 (#69)
| Benchmark | DeepSeek V4 Pro | Muse Spark |
|---|---|---|
| SciCode | 51% | 51.5% |
| LMArena Coding | 1470 | 1481 |
| SWE-bench Verified | 77.6% | — |
| FrontierCode | 28.6% | — |
| LMArena WebDev | 1582 | — |
| WeirdML | 66.2% | — |
| ALE-Bench | 1,403 | — |
Agentic & Tool Use Not comparable
DeepSeek V4 Pro: 32.8 (#58), Muse Spark: —
| Benchmark | DeepSeek V4 Pro | Muse Spark |
|---|---|---|
| APEX-Agents | 47.3% | — |
| Vending-Bench 2 | 3,285 | — |
Reasoning DeepSeek V4 Pro leads
DeepSeek V4 Pro: 56.5 (#24), Muse Spark: 35.9 (#67)
| Benchmark | DeepSeek V4 Pro | Muse Spark |
|---|---|---|
| CritPt | 18% | 11.3% |
| LMArena Hard Prompts | 1461 | 1474 |
| Epoch Capabilities Index | 155.31 | 152.04 |
| ARC-AGI-2 | 61.3% | — |
| Kagi LLM Benchmark | 53.5% | — |
| NYT Connections (extended) | 91.3% | — |
| ARC-AGI-1 | 90.5% | — |
| Chess Puzzles | 47% | — |
| Mystery Game Puzzles | 43% | — |
| DTBench | 93.9% | — |
| LMCA | 45.5% | — |
| Surface Evolver Bench | 40% | — |
| ForecastBench | 56.1 | — |
Math DeepSeek V4 Pro leads
DeepSeek V4 Pro: 64.8 (#30), Muse Spark: 47.8 (#66)
| Benchmark | DeepSeek V4 Pro | Muse Spark |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 98.6% | 88.9% |
| ProofBench | 50% | 17% |
| LMArena Math | 1455 | 1455 |
| FrontierMath (Tiers 1-3) | 64.6% | — |
| FrontierMath Tier 4 | 26.8% | — |
| MathArena Final-Answer Competitions | 76.6% | — |
| FrontierMath (Feb 2025 set) | — | 39% |
| FrontierMath Tier 4 (v1) | — | 14.6% |
Knowledge Muse Spark leads
DeepSeek V4 Pro: 59.5 (#31), Muse Spark: 65.7 (#13)
| Benchmark | DeepSeek V4 Pro | Muse Spark |
|---|---|---|
| GPQA Diamond | 91.7% | 89.8% |
| LMArena Expert | 1464 | 1457 |
| Humanity's Last Exam | — | 40.6% |
| SimpleQA Verified | 52.9% | — |
| Vectara Hallucination Rate | 8.6% | — |
Multimodal Not comparable
DeepSeek V4 Pro: —, Muse Spark: 43.4 (#24)
| Benchmark | DeepSeek V4 Pro | Muse Spark |
|---|---|---|
| LMArena Vision | — | 1306 |
| LMArena Document | — | 1444 |
Multilingual Muse Spark leads
DeepSeek V4 Pro: 54.4 (#45), Muse Spark: 56.1 (#24)
| Benchmark | DeepSeek V4 Pro | Muse Spark |
|---|---|---|
| LMArena Non-English | 1439 | 1464 |
| LMArena Chinese | 1486 | 1509 |
| LMArena French | 1472 | 1497 |
| LMArena German | 1458 | 1497 |
| LMArena Korean | 1447 | 1459 |
| LMArena Russian | 1453 | 1466 |
| LMArena Spanish | 1458 | 1472 |
| LMArena Japanese | 1445 | — |
Instruction Following Too close to call
DeepSeek V4 Pro: 76.1 (#47), Muse Spark: 75.9 (#51)
| Benchmark | DeepSeek V4 Pro | Muse Spark |
|---|---|---|
| LMArena Instruction Following | 1448 | 1442 |
Long Context Too close to call
DeepSeek V4 Pro: 45.0 (#51), Muse Spark: 44.4 (#69)
| Benchmark | DeepSeek V4 Pro | Muse Spark |
|---|---|---|
| LMArena Longer Query | 1458 | 1451 |
| CL-bench Life | 13.5% | — |
Writing & Preference Too close to call
DeepSeek V4 Pro: 65.5 (#46), Muse Spark: 66.0 (#39)
| Benchmark | DeepSeek V4 Pro | Muse Spark |
|---|---|---|
| LMArena Text | 1451 | 1474 |
| LMArena Creative Writing | 1446 | 1459 |
| LMArena Multi-Turn | 1467 | 1477 |
| EQ-Bench Creative Writing | 1553 | — |
| EQ-Bench 4 | 1166 | — |
Frequently asked questions
Is DeepSeek V4 Pro better than Muse Spark?
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 50.6 on the Noometry Index.
Is DeepSeek V4 Pro or Muse Spark better for coding?
DeepSeek V4 Pro scores higher on coding benchmarks: 52.4 versus 46.2 in the Noometry coding category.
How many benchmarks do DeepSeek V4 Pro and Muse Spark share?
22 benchmarks have published results for both models. DeepSeek V4 Pro has 48 scored results on Noometry and Muse Spark has 27.