Noometry

Model comparison

DeepSeek-V3.1-Terminus vs Grok 4

Grok 4 is the stronger model overall, scoring 48.1 to 43.1 on the Noometry Index.

Last verified . 11 shared benchmarks.

DeepSeek-V3.1-Terminus DeepSeek

43.1

Rank #97 Confirmed

Grok 4 xAI

48.1

Rank #56 Confirmed

Summary

  • They share 11 benchmarks with published results for both. DeepSeek-V3.1-Terminus scores higher in 2 categories and Grok 4 in 5 categories; 6 gaps are clear of the uncertainty.
  • The widest gap is in long context, where Grok 4 leads 63.1 to 43.4.
  • The biggest single-benchmark swing is Kagi LLM Benchmark: 57.4% for DeepSeek-V3.1-Terminus and 73.6% for Grok 4.
  • DeepSeek-V3.1-Terminus has downloadable open weights; the other is API-only.

Side by side

DeepSeek-V3.1-TerminusGrok 4
ProviderDeepSeekxAI
Noometry Index43.148.1
Released2025-09-222025-07-09
WeightsOpenProprietary
Context window164K—
Max output147K—
Input $ / M tokens$0.27—
Output $ / M tokens$1—
Results tracked1648

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

Category by category

Coding Grok 4 leads

DeepSeek-V3.1-Terminus: 42.0 (#113), Grok 4: 50.3 (#46)

BenchmarkDeepSeek-V3.1-TerminusGrok 4
LMArena Coding14261408
Aider Polyglot—79.6%
SciCode40.6%—
WeirdML—45.7%
ALE-Bench745.17—

Agentic & Tool Use Not comparable

DeepSeek-V3.1-Terminus: —, Grok 4: 32.3 (#68)

BenchmarkDeepSeek-V3.1-TerminusGrok 4
Terminal-Bench—27.2%
Berkeley Function Calling Leaderboard—63%
GDPval—21.1%
Cybench—43%
DeepResearch Bench—47.3%
BALROG—43.6%
LMArena Search—1142
METR Time Horizons—66.6%

Reasoning Grok 4 leads

DeepSeek-V3.1-Terminus: 26.4 (#133), Grok 4: 36.7 (#65)

BenchmarkDeepSeek-V3.1-TerminusGrok 4
Kagi LLM Benchmark57.4%73.6%
LMArena Hard Prompts14261409
ARC-AGI-2—16%
SimpleBench—60.5%
ARC-AGI-1—66.7%
CritPt1.7%—
Chess Puzzles—28%
DTBench81.3%—
LMCA28.6%—
Epoch Capabilities Index—146.44
ForecastBench—60.9

Math Grok 4 leads

DeepSeek-V3.1-Terminus: 38.5 (#137), Grok 4: 48.4 (#64)

BenchmarkDeepSeek-V3.1-TerminusGrok 4
LMArena Math14021422
OTIS Mock AIME 2024-2025—84%
Omni-MATH—60.3%
FrontierMath (Feb 2025 set)—19.7%
FrontierMath Tier 4 (v1)—2.1%

Knowledge Not comparable

DeepSeek-V3.1-Terminus: —, Grok 4: 53.8 (#55)

BenchmarkDeepSeek-V3.1-TerminusGrok 4
GPQA Diamond—87%
MMLU-Pro—85.1%
Confabulations—12.4%
GPQA (HELM)—72.7%
LMArena Expert—1415

Multimodal Not comparable

DeepSeek-V3.1-Terminus: —, Grok 4: 33.7 (#94)

BenchmarkDeepSeek-V3.1-TerminusGrok 4
LMArena Vision—1210
GeoBench—45%

Multilingual Too close to call

DeepSeek-V3.1-Terminus: 52.1 (#92), Grok 4: 51.8 (#103)

BenchmarkDeepSeek-V3.1-TerminusGrok 4
LMArena Non-English14071403
LMArena Russian14361410
LMArena Chinese—1427
LMArena French—1418
LMArena German—1429
LMArena Japanese—1394
LMArena Korean—1377
LMArena Spanish—1420

Instruction Following Grok 4 leads

DeepSeek-V3.1-Terminus: 74.0 (#106), Grok 4: 79.2 (#5)

BenchmarkDeepSeek-V3.1-TerminusGrok 4
LMArena Instruction Following14041387
IFEval—94.9%

Long Context Grok 4 leads

DeepSeek-V3.1-Terminus: 43.4 (#97), Grok 4: 63.1 (#4)

BenchmarkDeepSeek-V3.1-TerminusGrok 4
LMArena Longer Query14211409
Fiction.LiveBench—94.4%

Writing & Preference DeepSeek-V3.1-Terminus leads

DeepSeek-V3.1-Terminus: 61.0 (#92), Grok 4: 58.5 (#116)

BenchmarkDeepSeek-V3.1-TerminusGrok 4
LMArena Text14191411
LMArena Creative Writing14031397
LMArena Multi-Turn14111416
Short-Story Creative Writing—76.9%
WildBench—79.7%

Frequently asked questions

Is DeepSeek-V3.1-Terminus better than Grok 4?

Grok 4 is the stronger model overall, scoring 48.1 to 43.1 on the Noometry Index.

Is DeepSeek-V3.1-Terminus or Grok 4 better for coding?

Grok 4 scores higher on coding benchmarks: 50.3 versus 42.0 in the Noometry coding category.

How many benchmarks do DeepSeek-V3.1-Terminus and Grok 4 share?

11 benchmarks have published results for both models. DeepSeek-V3.1-Terminus has 16 scored results on Noometry and Grok 4 has 48.

Related comparisons

Go deeper