Noometry

Model comparison

DeepSeek V4.1 Flash vs Muse Spark

DeepSeek V4.1 Flash is the stronger model overall, scoring 52.8 to 50.6 on the Noometry Index.

Last verified . 23 shared benchmarks.

DeepSeek V4.1 Flash DeepSeek

52.8

Rank #38 Confirmed

Muse Spark Meta

50.6

Rank #46 Confirmed

Summary

  • They share 23 benchmarks with published results for both. DeepSeek V4.1 Flash scores higher in 5 categories and Muse Spark in 4 categories; 7 gaps are clear of the uncertainty.
  • The widest gap is in math, where DeepSeek V4.1 Flash leads 66.7 to 47.8.
  • The biggest single-benchmark swing is ProofBench: 54% for DeepSeek V4.1 Flash and 17% for Muse Spark.
  • DeepSeek V4.1 Flash has downloadable open weights; the other is API-only.

Side by side

DeepSeek V4.1 FlashMuse Spark
ProviderDeepSeekMeta
Noometry Index52.850.6
Released2026-09-092026-04-08
WeightsOpenProprietary
Context window1M—
Max output393K—
Input $ / M tokens$0.15—
Output $ / M tokens$0.60—
Results tracked3727

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding DeepSeek V4.1 Flash leads

DeepSeek V4.1 Flash: 52.9 (#32), Muse Spark: 46.2 (#69)

BenchmarkDeepSeek V4.1 FlashMuse Spark
SciCode51.9%51.5%
LMArena Coding15061481
LMArena WebDev1619—
ALE-Bench1,092—

Agentic & Tool Use Not comparable

DeepSeek V4.1 Flash: 31.2 (#69), Muse Spark: —

BenchmarkDeepSeek V4.1 FlashMuse Spark
APEX-Agents39.5%—
GDP.pdf19.8%—

Reasoning DeepSeek V4.1 Flash leads

DeepSeek V4.1 Flash: 50.2 (#36), Muse Spark: 35.9 (#67)

BenchmarkDeepSeek V4.1 FlashMuse Spark
CritPt14.3%11.3%
LMArena Hard Prompts14831474
Epoch Capabilities Index154.9152.04
NYT Connections (extended)89.6%—
Mystery Game Puzzles43%—
DTBench89.9%—
LMCA47%—
Surface Evolver Bench46.3%—

Math DeepSeek V4.1 Flash leads

DeepSeek V4.1 Flash: 66.7 (#25), Muse Spark: 47.8 (#66)

BenchmarkDeepSeek V4.1 FlashMuse Spark
OTIS Mock AIME 2024-202598.3%88.9%
ProofBench54%17%
LMArena Math14771455
FrontierMath (Tiers 1-3)67.4%—
FrontierMath Tier 426.8%—
FrontierMath (Feb 2025 set)—39%
FrontierMath Tier 4 (v1)—14.6%

Knowledge Muse Spark leads

DeepSeek V4.1 Flash: 57.9 (#38), Muse Spark: 65.7 (#13)

BenchmarkDeepSeek V4.1 FlashMuse Spark
GPQA Diamond89.8%89.8%
LMArena Expert15061457
Humanity's Last Exam—40.6%

Multimodal Muse Spark leads

DeepSeek V4.1 Flash: 39.1 (#61), Muse Spark: 43.4 (#24)

BenchmarkDeepSeek V4.1 FlashMuse Spark
LMArena Vision12771306
Furniture Assembly34.2%—
LMArena Document—1444

Multilingual Muse Spark leads

DeepSeek V4.1 Flash: 55.0 (#35), Muse Spark: 56.1 (#24)

BenchmarkDeepSeek V4.1 FlashMuse Spark
LMArena Non-English14481464
LMArena Chinese14971509
LMArena French14521497
LMArena German14841497
LMArena Korean14521459
LMArena Russian14711466
LMArena Spanish14591472
LMArena Japanese1412—

Instruction Following DeepSeek V4.1 Flash leads

DeepSeek V4.1 Flash: 77.3 (#26), Muse Spark: 75.9 (#51)

BenchmarkDeepSeek V4.1 FlashMuse Spark
LMArena Instruction Following14741442

Long Context Too close to call

DeepSeek V4.1 Flash: 45.2 (#47), Muse Spark: 44.4 (#69)

BenchmarkDeepSeek V4.1 FlashMuse Spark
LMArena Longer Query14751451

Writing & Preference Too close to call

DeepSeek V4.1 Flash: 65.4 (#48), Muse Spark: 66.0 (#39)

BenchmarkDeepSeek V4.1 FlashMuse Spark
LMArena Text14621474
LMArena Creative Writing14351459
LMArena Multi-Turn14571477
EQ-Bench Creative Writing1540—

Frequently asked questions

Is DeepSeek V4.1 Flash better than Muse Spark?

DeepSeek V4.1 Flash is the stronger model overall, scoring 52.8 to 50.6 on the Noometry Index.

Is DeepSeek V4.1 Flash or Muse Spark better for coding?

DeepSeek V4.1 Flash scores higher on coding benchmarks: 52.9 versus 46.2 in the Noometry coding category.

How many benchmarks do DeepSeek V4.1 Flash and Muse Spark share?

23 benchmarks have published results for both models. DeepSeek V4.1 Flash has 37 scored results on Noometry and Muse Spark has 27.

Related comparisons

Go deeper