Noometry

Model comparison

DeepSeek-V3.2-Exp vs MiniMax-M2.7

DeepSeek-V3.2-Exp is the stronger model overall, scoring 44.3 to 37.7 on the Noometry Index.

Last verified . 27 shared benchmarks.

DeepSeek-V3.2-Exp DeepSeek

44.3

Rank #78 Confirmed

MiniMax-M2.7 MiniMax

37.7

Rank #196 Confirmed

Summary

  • They share 27 benchmarks with published results for both. DeepSeek-V3.2-Exp scores higher in 9 categories and MiniMax-M2.7 in 0 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in math, where DeepSeek-V3.2-Exp leads 41.7 to 25.9.
  • The biggest single-benchmark swing is Thematic Generalization: 65% for DeepSeek-V3.2-Exp and 39.3% for MiniMax-M2.7.
  • DeepSeek-V3.2-Exp is cheaper at $0.26 / $0.38 per million input/output tokens, against $0.30 / $1.20 for MiniMax-M2.7.
  • MiniMax-M2.7 accepts more context: 205K tokens versus 164K.

Side by side

DeepSeek-V3.2-ExpMiniMax-M2.7
ProviderDeepSeekMiniMax
Noometry Index44.337.7
Released2025-09-292026-03-18
WeightsOpenOpen
Context window164K205K
Max output66K131K
Input $ / M tokens$0.26$0.30
Output $ / M tokens$0.38$1.20
Results tracked4930

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

Category by category

Coding DeepSeek-V3.2-Exp leads

DeepSeek-V3.2-Exp: 46.5 (#65), MiniMax-M2.7: 41.8 (#120)

BenchmarkDeepSeek-V3.2-ExpMiniMax-M2.7
LMArena WebDev13621398
SciCode38.9%47%
WeirdML39.5%37%
LMArena Coding14541454
SWE-bench Verified (bash only)70%—
Aider Polyglot74.2%—
SWE-bench Multilingual59%—
ALE-Bench—599.25

Agentic & Tool Use DeepSeek-V3.2-Exp leads

DeepSeek-V3.2-Exp: 32.7 (#59), MiniMax-M2.7: 25.1 (#111)

BenchmarkDeepSeek-V3.2-ExpMiniMax-M2.7
Terminal-Bench39.6%45.1%
APEX-Agents21.3%—
Berkeley Function Calling Leaderboard56.7%—
TheAgentCompany42.9%—
ExploitBench—13.3%
GBAEval—0%
Vending-Bench 21,034—

Reasoning DeepSeek-V3.2-Exp leads

DeepSeek-V3.2-Exp: 22.1 (#208), MiniMax-M2.7: 19.7 (#253)

BenchmarkDeepSeek-V3.2-ExpMiniMax-M2.7
NYT Connections (extended)36.7%24.7%
CritPt2.9%0.6%
Thematic Generalization65%39.3%
LMArena Hard Prompts14341422
Epoch Capabilities Index146.27145.85
ARC-AGI-24%—
Kagi LLM Benchmark52.2%—
ARC-AGI-157%—
Chess Puzzles14%—
DTBench87.7%—
LMCA29.1%—

Math DeepSeek-V3.2-Exp leads

DeepSeek-V3.2-Exp: 41.7 (#87), MiniMax-M2.7: 25.9 (#263)

BenchmarkDeepSeek-V3.2-ExpMiniMax-M2.7
ProofBench8%3%
LMArena Math14351420
MathArena Final-Answer Competitions57.7%—
OTIS Mock AIME 2024-202587.8%—
FrontierMath (Feb 2025 set)22.1%—
FrontierMath Tier 4 (v1)2.1%—

Knowledge DeepSeek-V3.2-Exp leads

DeepSeek-V3.2-Exp: 51.7 (#66), MiniMax-M2.7: 37.7 (#152)

BenchmarkDeepSeek-V3.2-ExpMiniMax-M2.7
Vectara Hallucination Rate5.3%12.9%
LMArena Expert14361444
GPQA Diamond83.4%—

Multilingual DeepSeek-V3.2-Exp leads

DeepSeek-V3.2-Exp: 52.2 (#90), MiniMax-M2.7: 50.3 (#123)

BenchmarkDeepSeek-V3.2-ExpMiniMax-M2.7
LMArena Non-English14091382
LMArena Chinese14611441
LMArena French14331421
LMArena German14401398
LMArena Japanese13741262
LMArena Korean13711313
LMArena Russian14241383
LMArena Spanish14401403

Instruction Following Too close to call

DeepSeek-V3.2-Exp: 74.5 (#93), MiniMax-M2.7: 74.1 (#103)

BenchmarkDeepSeek-V3.2-ExpMiniMax-M2.7
LMArena Instruction Following14131405

Long Context DeepSeek-V3.2-Exp leads

DeepSeek-V3.2-Exp: 47.6 (#16), MiniMax-M2.7: 43.3 (#99)

BenchmarkDeepSeek-V3.2-ExpMiniMax-M2.7
LMArena Longer Query14281419
Fiction.LiveBench83.3%—
CL-bench13.2%—
CL-bench Life9.5%—

Writing & Preference DeepSeek-V3.2-Exp leads

DeepSeek-V3.2-Exp: 62.4 (#77), MiniMax-M2.7: 58.9 (#112)

BenchmarkDeepSeek-V3.2-ExpMiniMax-M2.7
LMArena Text14251405
LMArena Creative Writing14031354
LMArena Multi-Turn14271412
EQ-Bench Creative Writing1515—

Frequently asked questions

Is DeepSeek-V3.2-Exp better than MiniMax-M2.7?

DeepSeek-V3.2-Exp is the stronger model overall, scoring 44.3 to 37.7 on the Noometry Index.

Which is cheaper, DeepSeek-V3.2-Exp or MiniMax-M2.7?

DeepSeek-V3.2-Exp is cheaper. It lists at $0.26 per million input tokens and $0.38 per million output tokens; MiniMax-M2.7 lists at $0.30 and $1.20.

Is DeepSeek-V3.2-Exp or MiniMax-M2.7 better for coding?

DeepSeek-V3.2-Exp scores higher on coding benchmarks: 46.5 versus 41.8 in the Noometry coding category.

Which has the bigger context window?

MiniMax-M2.7 does, with 205K tokens against 164K.

How many benchmarks do DeepSeek-V3.2-Exp and MiniMax-M2.7 share?

27 benchmarks have published results for both models. DeepSeek-V3.2-Exp has 49 scored results on Noometry and MiniMax-M2.7 has 30.

Related comparisons

Go deeper