Enter opens the highlighted result. Esc closes.

DeepSeek V3.2 Exp vs Kimi K2 Thinking

Kimi K2 Thinking leads DeepSeek V3.2 Exp on the Artificial Analysis Intelligence Index, 22.0 to 16.6, and DeepSeek V3.2 Exp costs 6.1x less per output token ($0.41 vs $2.50 per million). Updated October 22, 2026; 15 benchmarks have a result for both.

Replace DeepSeek V3.2 Exp Replace Kimi K2 Thinking
DeepSeek V3.2 ExpDeepSeek
Coding
No data
Overall
No data
Output $/M
$0.41
3
Kimi K2 ThinkingMoonshot AI
Coding
No data
Overall
No data
Output $/M
$2.50
11
vs boards won of 15 shared, 1 tied

DeepSeek V3.2 Exp or Kimi K2 Thinking: which to pick

Neither model has a coding score yet, so the two cannot be ranked on coding.

Pick DeepSeek V3.2 Exp for output tokens 6.1x cheaper ($0.41 vs $2.50 per million) and work measured by LiveCodeBench (Artificial Analysis run) or WeirdML, where it leads.

Model ADeepSeek V3.2 ExpDeepSeekNot benchmarked1 independent benchmarkNo cost per solved taskModel BKimi K2 ThinkingMoonshot AINot benchmarked1 independent benchmark$0.61 per solved task
Shared results
15
DeepSeek V3.2 Exp only
44
Kimi K2 Thinking only
15
Like-for-like categories
8 of 9

Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.

Which one for your work

Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.

At a glance

Coding index
DeepSeek V3.2 Exp not benchmarked, Kimi K2 Thinking not benchmarked.
Widest category
Agentic work, 59.1 points apart on average over 1 shared benchmark.
Widest single gap
τ²-Bench Telecom: DeepSeek V3.2 Exp 33.9% against Kimi K2 Thinking 93.0%, 59.1 points apart on a 0 to 100 scale.
Output price
DeepSeek V3.2 Exp $0.41 against Kimi K2 Thinking $2.50 per million output tokens, 6.1x apart.
Context window
DeepSeek V3.2 Exp takes 164K tokens, Kimi K2 Thinking 262K.

Benchmark by benchmark

First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.

Boards won, where both have a result
CategoryDeepSeek V3.2 ExpBarsKimi K2 ThinkingTies
Coding021 tie
Intelligence02
Agentic work01
Reasoning02
Math02
Knowledge10
Web design10
General12
All shared boards3111 tie
CodingLike-for-likeSix of these boards feed the Vibe Coding Index02
Coding benchmarks for DeepSeek V3.2 Exp and Kimi K2 Thinking
BenchmarkDeepSeek V3.2 ExpKimi K2 ThinkingLeader
Aider polyglot74.2%Not measuredOnly DeepSeek V3.2 Exp measured
SWE-bench VerifiedNot measured63.4%Only Kimi K2 Thinking measured
LMArena WebDev1272Not measuredOnly DeepSeek V3.2 Exp measured
LiveCodeBench (Vals AI)Not measured63.1%Only Kimi K2 Thinking measured
SWE-bench Verified (Vals AI)Not measured60.2%Only Kimi K2 Thinking measured
LiveCodeBench (Artificial Analysis run)78.9%85.3%Kimi K2 Thinking
Terminal-Bench Hard (Artificial Analysis run)31.1%31.1%Tie
Terminal-BenchNot measured35.7%Only Kimi K2 Thinking measured
WeirdML39.5%42.8%Kimi K2 Thinking
ALE-BenchNot measured597.5Only Kimi K2 Thinking measured
LMArena Text: coding1438Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: software and IT1438Not measuredOnly DeepSeek V3.2 Exp measured
LMArena WebDev: HTML1281Not measuredOnly DeepSeek V3.2 Exp measured
Neither model has a result on the other 26 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
Intelligence benchmarks for DeepSeek V3.2 Exp and Kimi K2 Thinking
BenchmarkDeepSeek V3.2 ExpKimi K2 ThinkingLeader
AA Intelligence Index16.622.0Kimi K2 Thinking
Epoch Capabilities Index144.9145.8Kimi K2 Thinking
Agentic workLike-for-likePublic boards, shown as published, kept out of the index01
Agentic work benchmarks for DeepSeek V3.2 Exp and Kimi K2 Thinking
BenchmarkDeepSeek V3.2 ExpKimi K2 ThinkingLeader
τ²-Bench Telecom33.9%93.0%Kimi K2 Thinking
METR time horizonsNot measured59.2%Only Kimi K2 Thinking measured
TheAgentCompany42.9%Not measuredOnly DeepSeek V3.2 Exp measured
Berkeley Function Calling Leaderboard (BFCL)56.7%Not measuredOnly DeepSeek V3.2 Exp measured
Neither model has a result on the other 37 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index02
Reasoning benchmarks for DeepSeek V3.2 Exp and Kimi K2 Thinking
BenchmarkDeepSeek V3.2 ExpKimi K2 ThinkingLeader
GPQA DiamondNot measured78.5%Only Kimi K2 Thinking measured
GPQA Diamond (Artificial Analysis run)79.7%83.8%Kimi K2 Thinking
Humanity's Last Exam (Artificial Analysis run)14.9%23.8%Kimi K2 Thinking
LMArena Text: hard prompts1429Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: expert1422Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: hard prompts in English1444Not measuredOnly DeepSeek V3.2 Exp measured
Kagi LLM Benchmark52.2%Not measuredOnly DeepSeek V3.2 Exp measured
Neither model has a result on the other 15 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index02
Math benchmarks for DeepSeek V3.2 Exp and Kimi K2 Thinking
BenchmarkDeepSeek V3.2 ExpKimi K2 ThinkingLeader
AIMENot measured85.4%Only Kimi K2 Thinking measured
MGSM (Vals AI)Not measured90.2%Only Kimi K2 Thinking measured
AA Math Index87.794.7Kimi K2 Thinking
AIME 2025 (Artificial Analysis run)87.7%94.7%Kimi K2 Thinking
LMArena Text: math1422Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: mathematical work1433Not measuredOnly DeepSeek V3.2 Exp measured
Neither model has a result on the other 7 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index10
Knowledge benchmarks for DeepSeek V3.2 Exp and Kimi K2 Thinking
BenchmarkDeepSeek V3.2 ExpKimi K2 ThinkingLeader
MMLU ProNot measured81.1%Only Kimi K2 Thinking measured
LegalBench (Vals AI)Not measured80.2%Only Kimi K2 Thinking measured
CaseLaw (Vals AI)Not measured65.7%Only Kimi K2 Thinking measured
CorpFin (Vals AI)Not measured60.6%Only Kimi K2 Thinking measured
TaxEval (Vals AI)Not measured71.7%Only Kimi K2 Thinking measured
MedQA (Vals AI)Not measured92.6%Only Kimi K2 Thinking measured
MMLU Pro (Artificial Analysis run)85.0%84.8%DeepSeek V3.2 Exp
LMArena Text: science1443Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: medicine and healthcare1468Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: legal and government1419Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: business and finance1412Not measuredOnly DeepSeek V3.2 Exp measured
Neither model has a result on the other 7 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index10
Web design benchmarks for DeepSeek V3.2 Exp and Kimi K2 Thinking
BenchmarkDeepSeek V3.2 ExpKimi K2 ThinkingLeader
Design Arena: websites11781113DeepSeek V3.2 Exp
Design Arena: UI components1164Not measuredOnly DeepSeek V3.2 Exp measured
Design Arena: game dev1144Not measuredOnly DeepSeek V3.2 Exp measured
Design Arena: data viz1152Not measuredOnly DeepSeek V3.2 Exp measured
Design Arena: 3D1148Not measuredOnly DeepSeek V3.2 Exp measured
Design Arena: all code categories1170Not measuredOnly DeepSeek V3.2 Exp measured
Design Arena: SVG1031Not measuredOnly DeepSeek V3.2 Exp measured
Neither model has a result on the other 2 boards here
GeneralLike-for-likePublic boards, shown as published, kept out of the index12
General benchmarks for DeepSeek V3.2 Exp and Kimi K2 Thinking
BenchmarkDeepSeek V3.2 ExpKimi K2 ThinkingLeader
IFBench54.1%68.1%Kimi K2 Thinking
Long context reasoning (AA-LCR)72.3%72.0%DeepSeek V3.2 Exp
DTBench87.7%Not measuredOnly DeepSeek V3.2 Exp measured
LMCA29.1%Not measuredOnly DeepSeek V3.2 Exp measured
Fiction.LiveBench83.3%Not measuredOnly DeepSeek V3.2 Exp measured
CL-bench13.2%17.6%Kimi K2 Thinking
CL-bench Life9.5%Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text1425Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: instruction following1401Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: creative writing1404Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: English1441Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: non-English1404Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: Chinese1461Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: German1440Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: Japanese1331Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: Korean1372Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: French1434Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: Spanish1441Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: Russian1412Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: Polish1435Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: multi-turn1422Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: longer queries1418Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: writing and language1407Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text: entertainment and media1407Not measuredOnly DeepSeek V3.2 Exp measured
Neither model has a result on the other 22 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for DeepSeek V3.2 Exp and Kimi K2 Thinking
BenchmarkDeepSeek V3.2 ExpKimi K2 ThinkingLeader
Neither model has a result on 5 boards here

What each workload costs

Three fixed token mixes at each model's listed price, so per-token rates become a bill.

Chat turn

1K input + 500 output tokens

DeepSeek V3.2 Exp$0.0005Fits in one request
Kimi K2 Thinking$0.0019Fits in one request

DeepSeek V3.2 Exp costs less

Repository review

50K input + 3K output tokens

DeepSeek V3.2 Exp$0.015Fits in one request
Kimi K2 Thinking$0.037Fits in one request

DeepSeek V3.2 Exp costs less

Agent loop

200K input + 10K output tokens

DeepSeek V3.2 Exp$0.058Exceeds the context window
Kimi K2 Thinking$0.145Fits in one request

DeepSeek V3.2 Exp costs less

Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.

Specification differences

The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.

Specifications of DeepSeek V3.2 Exp and Kimi K2 Thinking
SpecificationDeepSeek V3.2 ExpMoonshotAI Kimi K2 Thinking
MakerDeepSeekMoonshot AI
WeightsOpen weightsOpen weights
ReleasedSeptember 29, 2025November 6, 2025
Context window164K262K
Input $/Mtok$0.27$0.60
Output $/Mtok$0.41$2.50
Coding indexNot benchmarkedNot benchmarked
Independent measurements behind it11
Cost per solved taskNot computable$0.61
SourceOpenRouter listingOpenRouter listing

Questions

Is DeepSeek V3.2 Exp better than Kimi K2 Thinking for coding?

Neither model has a coding score yet, so the two cannot be ranked on coding. Pick DeepSeek V3.2 Exp for output tokens 6.1x cheaper ($0.41 vs $2.50 per million) and work measured by LiveCodeBench (Artificial Analysis run) or WeirdML, where it leads.

Which is cheaper, DeepSeek V3.2 Exp or Kimi K2 Thinking?

DeepSeek V3.2 Exp costs $0.27 per million input tokens and $0.41 per million output tokens; Kimi K2 Thinking costs $0.60 and $2.50. Cost per solved task is known for Kimi K2 Thinking only.

Which has the bigger context window, DeepSeek V3.2 Exp or Kimi K2 Thinking?

Kimi K2 Thinking, with 262K tokens against 164K.

How many benchmarks is the DeepSeek V3.2 Exp vs Kimi K2 Thinking comparison based on?

15 benchmarks have a result for both models as of October 22, 2026. DeepSeek V3.2 Exp's coding index rests on 1 independent benchmark and Kimi K2 Thinking's on 1. Where only one model has a result, the page says so instead of naming a winner.

Compare next