Noometry

Model comparison

DeepSeek V4 Pro vs Llama 4 Maverick

DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 30.9 on the Noometry Index. Llama 4 Maverick costs 3.3× less per token, which makes it the better buy when DeepSeek V4 Pro's lead doesn't matter for your workload.

Last verified . 33 shared benchmarks.

DeepSeek V4 Pro DeepSeek

54.3

Rank #31 Confirmed

Llama 4 Maverick Meta

30.9

Rank #282 Confirmed

Summary

  • They share 33 benchmarks with published results for both. DeepSeek V4 Pro scores higher in 9 categories and Llama 4 Maverick in 0 categories; 9 gaps are clear of the uncertainty.
  • The widest gap is in reasoning, where DeepSeek V4 Pro leads 56.5 to 10.1.
  • The biggest single-benchmark swing is ARC-AGI-1: 90.5% for DeepSeek V4 Pro and 4.4% for Llama 4 Maverick.
  • Llama 4 Maverick is cheaper at $0.19 / $0.65 per million input/output tokens, against $0.66 / $1.98 for DeepSeek V4 Pro.
  • DeepSeek V4 Pro accepts more context: 1M tokens versus 128K.

Side by side

DeepSeek V4 ProLlama 4 Maverick
ProviderDeepSeekMeta
Noometry Index54.330.9
Released2026-04-242025-04-05
WeightsOpenOpen
Context window1M128K
Max output393K4K
Input $ / M tokens$0.66$0.19
Output $ / M tokens$1.98$0.65
Results tracked4854

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), Llama 4 Maverick: 26.6 (#324)

BenchmarkDeepSeek V4 ProLlama 4 Maverick
SciCode51%33.1%
WeirdML66.2%24.5%
LMArena Coding14701302
ALE-Bench1,403172.97
SWE-bench Verified77.6%—
FrontierCode28.6%—
SWE-bench Verified (bash only)—21%
Aider Polyglot—15.6%
LMArena WebDev1582—
BigCodeBench Instruct—49.7%
BigCodeBench Complete—61.4%

Agentic & Tool Use DeepSeek V4 Pro leads

DeepSeek V4 Pro: 32.8 (#58), Llama 4 Maverick: 28.2 (#91)

BenchmarkDeepSeek V4 ProLlama 4 Maverick
APEX-Agents47.3%—
Berkeley Function Calling Leaderboard—37.3%
Vending-Bench 23,285—

Reasoning DeepSeek V4 Pro leads

DeepSeek V4 Pro: 56.5 (#24), Llama 4 Maverick: 10.1 (#342)

BenchmarkDeepSeek V4 ProLlama 4 Maverick
ARC-AGI-261.3%0%
Kagi LLM Benchmark53.5%55.9%
NYT Connections (extended)91.3%8%
ARC-AGI-190.5%4.4%
CritPt18%0%
LMArena Hard Prompts14611281
DTBench93.9%61.9%
LMCA45.5%15.9%
Epoch Capabilities Index155.31132.2
ForecastBench56.157.5
SimpleBench—27.7%
Chess Puzzles47%—
EnigmaEval—0.6%
Mystery Game Puzzles43%—
Surface Evolver Bench40%—

Math DeepSeek V4 Pro leads

DeepSeek V4 Pro: 64.8 (#30), Llama 4 Maverick: 26.0 (#262)

BenchmarkDeepSeek V4 ProLlama 4 Maverick
OTIS Mock AIME 2024-202598.6%20.6%
LMArena Math14551299
FrontierMath (Tiers 1-3)64.6%—
FrontierMath Tier 426.8%—
MathArena Final-Answer Competitions76.6%—
ProofBench50%—
Omni-MATH—42.2%
MATH Level 5—73%
FrontierMath (Feb 2025 set)—0.7%

Knowledge DeepSeek V4 Pro leads

DeepSeek V4 Pro: 59.5 (#31), Llama 4 Maverick: 33.4 (#204)

BenchmarkDeepSeek V4 ProLlama 4 Maverick
GPQA Diamond91.7%67%
Vectara Hallucination Rate8.6%8.2%
LMArena Expert14641259
Humanity's Last Exam—5.7%
SimpleQA Verified52.9%—
MMLU-Pro—81%
Confabulations—22.6%
GPQA (HELM)—65%

Multimodal Not comparable

DeepSeek V4 Pro: —, Llama 4 Maverick: 31.6 (#105)

BenchmarkDeepSeek V4 ProLlama 4 Maverick
LMArena Vision—1142
GeoBench—52%
SpatialViz-Bench—31.8%

Multilingual DeepSeek V4 Pro leads

DeepSeek V4 Pro: 54.4 (#45), Llama 4 Maverick: 42.2 (#195)

BenchmarkDeepSeek V4 ProLlama 4 Maverick
LMArena Non-English14391269
LMArena Chinese14861277
LMArena French14721259
LMArena German14581291
LMArena Japanese14451207
LMArena Korean14471203
LMArena Russian14531286
LMArena Spanish14581293

Instruction Following DeepSeek V4 Pro leads

DeepSeek V4 Pro: 76.1 (#47), Llama 4 Maverick: 71.7 (#146)

BenchmarkDeepSeek V4 ProLlama 4 Maverick
LMArena Instruction Following14481267
IFEval—90.8%

Long Context DeepSeek V4 Pro leads

DeepSeek V4 Pro: 45.0 (#51), Llama 4 Maverick: 31.4 (#279)

BenchmarkDeepSeek V4 ProLlama 4 Maverick
LMArena Longer Query14581280
Fiction.LiveBench—46.2%
CL-bench Life13.5%—

Writing & Preference DeepSeek V4 Pro leads

DeepSeek V4 Pro: 65.5 (#46), Llama 4 Maverick: 38.8 (#252)

BenchmarkDeepSeek V4 ProLlama 4 Maverick
LMArena Text14511287
LMArena Creative Writing14461267
EQ-Bench Creative Writing1553860
LMArena Multi-Turn14671289
Short-Story Creative Writing—62%
WildBench—80%
EQ-Bench 41166—

Frequently asked questions

Is DeepSeek V4 Pro better than Llama 4 Maverick?

DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 30.9 on the Noometry Index. Llama 4 Maverick costs 3.3× less per token, which makes it the better buy when DeepSeek V4 Pro's lead doesn't matter for your workload.

Which is cheaper, DeepSeek V4 Pro or Llama 4 Maverick?

Llama 4 Maverick is cheaper. It lists at $0.19 per million input tokens and $0.65 per million output tokens; DeepSeek V4 Pro lists at $0.66 and $1.98.

Is DeepSeek V4 Pro or Llama 4 Maverick better for coding?

DeepSeek V4 Pro scores higher on coding benchmarks: 52.4 versus 26.6 in the Noometry coding category.

Which has the bigger context window?

DeepSeek V4 Pro does, with 1M tokens against 128K.

How many benchmarks do DeepSeek V4 Pro and Llama 4 Maverick share?

33 benchmarks have published results for both models. DeepSeek V4 Pro has 48 scored results on Noometry and Llama 4 Maverick has 54.

Related comparisons

Go deeper