Noometry

Model comparison

DeepSeek V4 Flash vs Tulu 3 (Tülu 3) 70B

DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 32.6 on the Noometry Index.

Last verified . 13 shared benchmarks.

DeepSeek V4 Flash DeepSeek

53.6

Rank #32 Confirmed

Tulu 3 (Tülu 3) 70B Allen Institute for AI (Ai2)

32.6

Rank #242 Confirmed

Summary

  • They share 13 benchmarks with published results for both. DeepSeek V4 Flash scores higher in 8 categories and Tulu 3 (Tülu 3) 70B in 0 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in math, where DeepSeek V4 Flash leads 60.0 to 13.3.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 94.4% for DeepSeek V4 Flash and 4.4% for Tulu 3 (Tülu 3) 70B.

Side by side

DeepSeek V4 FlashTulu 3 (Tülu 3) 70B
ProviderDeepSeekAllen Institute for AI (Ai2)
Noometry Index53.632.6
Released2026-04-242024-11-21
WeightsOpenOpen
Context window1M—
Max output393K—
Input $ / M tokens$0.15—
Output $ / M tokens$0.60—
Results tracked4114

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

Category by category

Coding DeepSeek V4 Flash leads

DeepSeek V4 Flash: 47.6 (#56), Tulu 3 (Tülu 3) 70B: 35.4 (#211)

BenchmarkDeepSeek V4 FlashTulu 3 (Tülu 3) 70B
LMArena Coding14571236
FrontierCode18.8%—
LMArena WebDev1581—
SciCode49.9%—
WeirdML63%—
ALE-Bench1,306—

Reasoning DeepSeek V4 Flash leads

DeepSeek V4 Flash: 53.7 (#25), Tulu 3 (Tülu 3) 70B: 22.9 (#172)

BenchmarkDeepSeek V4 FlashTulu 3 (Tülu 3) 70B
LMArena Hard Prompts14431220
ARC-AGI-261.4%—
SimpleBench61.1%—
Kagi LLM Benchmark52.2%—
NYT Connections (extended)89.6%—
ARC-AGI-189%—
CritPt16.6%—
Chess Puzzles33%—
Mystery Game Puzzles34%—
DTBench90.9%—
LMCA41.7%—
Epoch Capabilities Index154.51—

Math DeepSeek V4 Flash leads

DeepSeek V4 Flash: 60.0 (#31), Tulu 3 (Tülu 3) 70B: 13.3 (#297)

BenchmarkDeepSeek V4 FlashTulu 3 (Tülu 3) 70B
OTIS Mock AIME 2024-202594.4%4.4%
LMArena Math14271242
FrontierMath (Tiers 1-3)57.5%—
FrontierMath Tier 424.4%—
MathArena Final-Answer Competitions76.5%—
ProofBench56%—
MATH Level 5—42.7%

Knowledge DeepSeek V4 Flash leads

DeepSeek V4 Flash: 55.5 (#43), Tulu 3 (Tülu 3) 70B: 25.1 (#257)

BenchmarkDeepSeek V4 FlashTulu 3 (Tülu 3) 70B
GPQA Diamond91%46.3%
SimpleQA Verified33.6%—
LMArena Expert1441—

Multilingual DeepSeek V4 Flash leads

DeepSeek V4 Flash: 53.2 (#71), Tulu 3 (Tülu 3) 70B: 40.1 (#217)

BenchmarkDeepSeek V4 FlashTulu 3 (Tülu 3) 70B
LMArena Non-English14201236
LMArena Chinese14671249
LMArena Russian14281246
LMArena French1438—
LMArena German1418—
LMArena Japanese1403—
LMArena Korean1383—
LMArena Spanish1437—

Instruction Following DeepSeek V4 Flash leads

DeepSeek V4 Flash: 75.1 (#76), Tulu 3 (Tülu 3) 70B: 65.0 (#222)

BenchmarkDeepSeek V4 FlashTulu 3 (Tülu 3) 70B
LMArena Instruction Following14211233

Long Context DeepSeek V4 Flash leads

DeepSeek V4 Flash: 43.8 (#81), Tulu 3 (Tülu 3) 70B: 37.1 (#217)

BenchmarkDeepSeek V4 FlashTulu 3 (Tülu 3) 70B
LMArena Longer Query14351224

Writing & Preference DeepSeek V4 Flash leads

DeepSeek V4 Flash: 64.0 (#59), Tulu 3 (Tülu 3) 70B: 45.8 (#218)

BenchmarkDeepSeek V4 FlashTulu 3 (Tülu 3) 70B
LMArena Text14321256
LMArena Creative Writing14031231
LMArena Multi-Turn14491252
EQ-Bench Creative Writing1559—

Frequently asked questions

Is DeepSeek V4 Flash better than Tulu 3 (Tülu 3) 70B?

DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 32.6 on the Noometry Index.

Is DeepSeek V4 Flash or Tulu 3 (Tülu 3) 70B better for coding?

DeepSeek V4 Flash scores higher on coding benchmarks: 47.6 versus 35.4 in the Noometry coding category.

How many benchmarks do DeepSeek V4 Flash and Tulu 3 (Tülu 3) 70B share?

13 benchmarks have published results for both models. DeepSeek V4 Flash has 41 scored results on Noometry and Tulu 3 (Tülu 3) 70B has 14.

Related comparisons

Go deeper