Skip to content
VECTOR WIREAI INTELLIGENCE
⌘K
UTC
Models741Deals30d271Policy30d66, 14 pendingSources1,015

DeepSeek-V3.2 vs Kimi K2 Thinking

DeepSeek · released

Wins 2 of 7 areas

Coding · Long documents

Moonshot · released

Wins 1 of 7 areas

Following instructions

DeepSeek-V3.2 wins more areas, narrowly.Kimi K2 Thinking is better at following instructions.

Scores updated · 50 tests both models report · How we compare

Where each one wins

Tests won in each of the seven areas where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.

Math and following instructions rest on a single test each.

What it costs

Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.

DeepSeek-V3.2$0.28 to read · $0.42 to write$0.70Price from Artificial Analysis
Kimi K2 Thinking$0.60 to read · $2.50 to write$3.10Price from Artificial Analysis

DeepSeek-V3.2 costs 77% less for the same work.

The biggest differences

The three tests each model wins by the widest margin. Scores are out of 100.

Where DeepSeek-V3.2 pulls ahead

  • Questions about very long textsLongBench v2+14.7points aheadDeepSeek-V3.259.8Kimi K2 Thinking45.1
  • Finds hard-to-locate facts by browsing the webBrowseComp+9.9points aheadDeepSeek-V3.251.4Kimi K2 Thinking41.5
  • Fixes real GitHub issues in Python projectsSWE-bench Verified+9.7points aheadDeepSeek-V3.273.1Kimi K2 Thinking63.4

Where Kimi K2 Thinking pulls ahead

  • Real work tasks from 44 professionsGDPVal+14.7points aheadKimi K2 Thinking24.5DeepSeek-V3.29.8
  • Follows unfamiliar, precisely checkable instructionsIFBench+7.4points aheadKimi K2 Thinking68.1DeepSeek-V3.260.7
  • Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+7.1points aheadKimi K2 Thinking24.4DeepSeek-V3.217.3

Every test, side by side

All 50 tests both models report. The winning score is in its model's colour; marks a score checked independently.

CodingDeepSeek-V3.2

Full coding ranking

AgentsEven
  • GDPValKimi K2 Thinking by 14.79.824.5+14.7
  • BrowseCompDeepSeek-V3.2 by 9.951.441.5+9.9

Full agents ranking

ReasoningEven

Full reasoning ranking

FactsEven

Full facts ranking

MathEven

Full math ranking

Long documentsDeepSeek-V3.2

Full long documents ranking

Following instructionsKimi K2 Thinking
  • IFBenchKimi K2 Thinking by 7.460.768.1+7.4

Full following instructions ranking

Other results34 tests, not counted

Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.

Questions people ask

Which is better, DeepSeek-V3.2 or Kimi K2 Thinking?

DeepSeek-V3.2 wins two of the seven areas where both have results: coding and long documents. Kimi K2 Thinking wins following instructions. They are level on agents, reasoning, facts and math.

Which is better for coding?

DeepSeek-V3.2. It wins 3 of the 5 coding tests both models report; Kimi K2 Thinking wins 1, and 1 is a tie.

Which is cheaper?

DeepSeek-V3.2 costs $0.28 per million input tokens and $0.42 per million output tokens; Kimi K2 Thinking costs $0.60 and $2.50. That makes DeepSeek-V3.2 about 77% cheaper for the same work.

How do you compare the two?

We use the 50 benchmark tests both models have published scores on. The verdict counts the 16 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 34 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.

Compare DeepSeek-V3.2 with

DeepSeek

Across the field

All comparisons