Noometry

Model comparison

DeepSeek V4 Flash vs Grok 3

DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 39.9 on the Noometry Index.

Last verified . 26 shared benchmarks.

DeepSeek V4 Flash DeepSeek

53.6

Rank #35 Confirmed

Grok 3 xAI

39.9

Rank #157 Confirmed

Summary

  • They share 26 benchmarks with published results for both. DeepSeek V4 Flash scores higher in 7 categories and Grok 3 in 1 category; 6 gaps are clear of the uncertainty.
  • The widest gap is in reasoning, where DeepSeek V4 Flash leads 53.7 to 13.7.
  • The biggest single-benchmark swing is ARC-AGI-1: 89% for DeepSeek V4 Flash and 5.5% for Grok 3.
  • DeepSeek V4 Flash has downloadable open weights; the other is API-only.

Side by side

DeepSeek V4 FlashGrok 3
ProviderDeepSeekxAI
Noometry Index53.639.9
Released2026-04-242025-04-09
WeightsOpenProprietary
Context window1M—
Max output393K—
Input $ / M tokens$0.15—
Output $ / M tokens$0.60—
Results tracked4140

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

Category by category

Coding DeepSeek V4 Flash leads

DeepSeek V4 Flash: 47.9 (#59), Grok 3: 41.9 (#115)

BenchmarkDeepSeek V4 FlashGrok 3
WeirdML63%37.2%
LMArena Coding14571432
FrontierCode18.8%—
Aider Polyglot—53.3%
LMArena WebDev1582—
SciCode49.9%—
ALE-Bench1,306—

Agentic & Tool Use Not comparable

DeepSeek V4 Flash: —, Grok 3: 30.5 (#76)

BenchmarkDeepSeek V4 FlashGrok 3
BALROG—29.5%

Reasoning DeepSeek V4 Flash leads

DeepSeek V4 Flash: 53.7 (#30), Grok 3: 13.7 (#333)

BenchmarkDeepSeek V4 FlashGrok 3
ARC-AGI-261.4%0%
SimpleBench61.1%36.1%
Kagi LLM Benchmark52.2%61.3%
ARC-AGI-189%5.5%
LMArena Hard Prompts14441434
Epoch Capabilities Index154.49138.33
NYT Connections (extended)89.6%—
CritPt16.6%—
Chess Puzzles33%—
Mystery Game Puzzles34%—
DTBench90.9%—
LMCA41.7%—

Math DeepSeek V4 Flash leads

DeepSeek V4 Flash: 60.3 (#37), Grok 3: 38.0 (#145)

BenchmarkDeepSeek V4 FlashGrok 3
OTIS Mock AIME 2024-202594.4%55.6%
LMArena Math14271391
FrontierMath (Tiers 1-3)57.5%—
FrontierMath Tier 424.4%—
MathArena Final-Answer Competitions76.5%—
ProofBench56%—
Omni-MATH—46.4%
MATH Level 5—88.7%
FrontierMath (Feb 2025 set)—3.8%
FrontierMath Tier 4 (v1)—0%

Knowledge DeepSeek V4 Flash leads

DeepSeek V4 Flash: 55.4 (#48), Grok 3: 46.2 (#82)

BenchmarkDeepSeek V4 FlashGrok 3
GPQA Diamond91%75.8%
LMArena Expert14411421
SimpleQA Verified33.6%—
MMLU-Pro—78.8%
Confabulations—14.2%
Vectara Hallucination Rate—5.8%
GPQA (HELM)—65%

Multilingual Too close to call

DeepSeek V4 Flash: 53.0 (#72), Grok 3: 52.3 (#87)

BenchmarkDeepSeek V4 FlashGrok 3
LMArena Non-English14201410
LMArena Chinese14681448
LMArena French14391460
LMArena German14181431
LMArena Japanese14061387
LMArena Korean13841373
LMArena Russian14281416
LMArena Spanish14361417

Instruction Following Too close to call

DeepSeek V4 Flash: 74.9 (#81), Grok 3: 75.0 (#73)

BenchmarkDeepSeek V4 FlashGrok 3
LMArena Instruction Following14211409
IFEval—88.4%

Long Context DeepSeek V4 Flash leads

DeepSeek V4 Flash: 43.8 (#85), Grok 3: 38.7 (#192)

BenchmarkDeepSeek V4 FlashGrok 3
LMArena Longer Query14341439
Fiction.LiveBench—58.3%

Writing & Preference DeepSeek V4 Flash leads

DeepSeek V4 Flash: 63.8 (#61), Grok 3: 55.8 (#141)

BenchmarkDeepSeek V4 FlashGrok 3
LMArena Text14321426
LMArena Creative Writing14031414
EQ-Bench Creative Writing15591186
LMArena Multi-Turn14491425
Short-Story Creative Writing—76.4%
WildBench—84.9%

Frequently asked questions

Is DeepSeek V4 Flash better than Grok 3?

DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 39.9 on the Noometry Index.

Is DeepSeek V4 Flash or Grok 3 better for coding?

DeepSeek V4 Flash scores higher on coding benchmarks: 47.9 versus 41.9 in the Noometry coding category.

How many benchmarks do DeepSeek V4 Flash and Grok 3 share?

26 benchmarks have published results for both models. DeepSeek V4 Flash has 41 scored results on Noometry and Grok 3 has 40.

Related comparisons

Go deeper