Skip to main content
BenchLM
Advertise

Model comparison

DeepSeek V4 Pro vs Kimi K2.5

Data verified

Head-to-head evidence from 15 shared benchmark results across 5 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.

60.66/100
Margin
1.0pts
← winning
Kimi K2.5
Moonshot AI
59.66/100
2 category wins2 category wins

Public leaderboard positions: DeepSeek V4 Pro #46 (Supported); Kimi K2.5 #54 (Supported). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.

Evidence parity. DeepSeek V4 Pro and Kimi K2.5 share 15 comparable benchmark results. 4 of 8 categories are comparable. 8 results are unique to DeepSeek V4 Pro; 48 to Kimi K2.5.

Updated July 23, 2026
Shared results
15
DeepSeek V4 Pro only
8
Kimi K2.5 only
48
Comparable categories
4 / 8

Pick DeepSeek V4 Pro if you want the stronger benchmark profile. Kimi K2.5 only becomes the better choice if mathematics is the priority.

Confidence note. This is a partial-evidence comparison with 15 shared benchmark results across 5 evidence categories; 4 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.

Why this result

DeepSeek V4 Pro finishes one point ahead on BenchLM's BenchAlign leaderboard, 60.66 to 59.66. That is enough to call, but not enough to treat as a blowout. This matchup comes down to a few meaningful edges rather than one model dominating the board.

DeepSeek V4 Pro's sharpest advantage is in coding, where it averages 65.3 against 59.4. The single biggest benchmark swing on the page is HMMT Feb 2026, 31.7% to 87.1%. Kimi K2.5 does hit back in mathematics, so the answer changes if that is the part of the workload you care about most.

Kimi K2.5 is also the more expensive model on tokens at $0.60 input / $3.00 output per 1M tokens, versus $0.43 input / $0.87 output per 1M tokens for DeepSeek V4 Pro. That is roughly 3.4x on output cost alone. DeepSeek V4 Pro gives you the larger context window at 1M, compared with 256K for Kimi K2.5.

Category breakdown

Exact category averages are shown below. Not measured means BenchLM does not have enough sourced public coverage for that model and category.

Category scores and score margins for DeepSeek V4 Pro and Kimi K2.5
CategoryDeepSeek V4 ProΔKimi K2.5
MathDeepSeek V4 Pro31.7Margin→ 28.9Kimi K2.560.6
KnowledgeDeepSeek V4 Pro41.3Margin→ 15.6Kimi K2.556.9
CodingDeepSeek V4 Pro65.3Margin← 5.9Kimi K2.559.4
AgenticDeepSeek V4 Pro59.1Margin← 4.1Kimi K2.555.0
ReasoningDeepSeek V4 ProNot measuredMarginNo overlapKimi K2.561.0
MultilingualDeepSeek V4 ProNot measuredMarginNo overlapKimi K2.582.3
MultimodalDeepSeek V4 ProNot measuredMarginNo overlapKimi K2.578.5
Inst. FollowingDeepSeek V4 ProNot measuredMarginNo overlapKimi K2.593.9

Decisive benchmark drivers

The largest measured benchmark gaps in this matchup, with exact reported values.

More
A · DeepSeek V4 ProB · Kimi K2.5

  1. HMMT Feb 2026

    Math
    Source ↗
    A 31.7%B 87.1%
    Winner: Kimi K2.5Δ 55.4
    HMMT Feb 2026: DeepSeek V4 Pro scored 31.7%; Kimi K2.5 scored 87.1%. Kimi K2.5 wins this benchmark.

  2. HLE

    Knowledge
    Source ↗
    A 7.7%B 30.1%
    Winner: Kimi K2.5Δ 22.4
    HLE: DeepSeek V4 Pro scored 7.7%; Kimi K2.5 scored 30.1%. Kimi K2.5 wins this benchmark.

  3. GPQA

    Knowledge
    Source ↗
    A 72.9%B 87.6%
    Winner: Kimi K2.5Δ 14.7
    GPQA: DeepSeek V4 Pro scored 72.9%; Kimi K2.5 scored 87.6%. Kimi K2.5 wins this benchmark.

  4. Terminal-Bench 2.0

    Agentic
    Source ↗
    A 59.1%B 50.8%
    Winner: DeepSeek V4 ProΔ 8.3
    Terminal-Bench 2.0: DeepSeek V4 Pro scored 59.1%; Kimi K2.5 scored 50.8%. DeepSeek V4 Pro wins this benchmark.

  5. MMLU-Pro

    Knowledge
    Source ↗
    A 82.9%B 87.1%
    Winner: Kimi K2.5Δ 4.2
    MMLU-Pro: DeepSeek V4 Pro scored 82.9%; Kimi K2.5 scored 87.1%. Kimi K2.5 wins this benchmark.

Operational comparison

Runtime and commercial metrics are compared only when both models have a complete sourced value.

MetricDeepSeek V4 ProKimi K2.5Comparison
Input / output priceUSD per 1M tokensDeepSeek V4 Pro$0.435 input / $0.87 outputKimi K2.5$0.6 input / $3 outputDeepSeek V4 Pro has the lower combined listed price.
Generation speedtokens per secondDeepSeek V4 ProNot availableKimi K2.545 tok/sA complete speed comparison is not available.
First-answer latencyseconds to first tokenDeepSeek V4 ProNot availableKimi K2.52.38 sA complete latency comparison is not available.
Context windowmaximum listed tokensDeepSeek V4 Pro1MKimi K2.5256KDeepSeek V4 Pro lists the larger context window.

Benchmark Deep Dive

AgenticDeepSeek V4 Pro wins
BenchmarkDeepSeek V4 ProKimi K2.5Result
Terminal-Bench 2.0Source 59.1%50.8%DeepSeek V4 Pro leads
MCP AtlasSource 69.4%29.5%DeepSeek V4 Pro leads
ToolathlonSource 46.3%27.8%DeepSeek V4 Pro leads
Claw-EvalSource 59.8%52.3%DeepSeek V4 Pro leads
Gert LabsSource 50.28%45.88%DeepSeek V4 Pro leads
ResearchClawBenchSource 17.1%14.0%DeepSeek V4 Pro leads
BrowseCompSource —60.6%Not comparable
QwenClawBenchSource —54.3%Not comparable
τ³-bench resultsSource —65.7%Not comparable
DeepSearchQASource —77.1%Not comparable
DeepPlanningSource —14.4%Not comparable
MCP-TasksSource —59.1%Not comparable
WideResearchSource —72.7%Not comparable
τ²-bench resultsSource —95.9%Not comparable
APEX-Agents-AASource —11.5%Not comparable
JobBenchSource —8.7%Not comparable
AA Agentic IndexSource —21.7%Not comparable
GDPval-AASource —25.4%Not comparable
GDPval-AASource —1009Not comparable
CodingDeepSeek V4 Pro wins
BenchmarkDeepSeek V4 ProKimi K2.5Result
SWE-bench VerifiedSource 73.6%76.8%Kimi K2.5 leads
SWE-bench ProSource 52.1%50.7%DeepSeek V4 Pro leads
SWE MultilingualSource 69.8%73%Kimi K2.5 leads
Terminal-Bench 2.0Source 59.1%—Not comparable
SWE-bench Verified*Source —70.8%Not comparable
LiveCodeBench v6Source —85.0%Not comparable
SWE-RebenchSource —58.5%Not comparable
React Native EvalsSource —77.2%Not comparable
SciCodeSource —48.7%Not comparable
AA-SciCodeSource —49.0%Not comparable
AA Coding IndexSource —46.8%Not comparable
Reasoning
BenchmarkDeepSeek V4 ProKimi K2.5Result
MRCR 1MSource 44.7%—Not comparable
CorpusQA 1MSource 35.6%—Not comparable
LongBench v2Source —61%Not comparable
AA-LCRSource —65.3%Not comparable
CritPtSource —3.1%Not comparable
KnowledgeKimi K2.5 wins
BenchmarkDeepSeek V4 ProKimi K2.5Result
MMLU-ProSource 82.9%87.1%Kimi K2.5 leads
SimpleQASource 45%—Not comparable
Chinese-SimpleQASource 75.8%—Not comparable
GPQASource 72.9%87.6%Kimi K2.5 leads
GPQA-DSource 72.9%87.6%Kimi K2.5 leads
HLESource 7.7%30.1%Kimi K2.5 leads
SuperGPQASource —69.2%Not comparable
MMLU-Pro (Arcee)Source —87.1%Not comparable
Artificial Analysis Intelligence IndexSource —35.4%Not comparable
AA-GPQA DiamondSource —87.9%Not comparable
AA-HLESource —29.4%Not comparable
AA-Omniscience IndexSource —-8.1%Not comparable
AA-Omniscience AccuracySource —34.3%Not comparable
AA-Omniscience Hallucination RateSource —64.6%Not comparable
MathKimi K2.5 wins
BenchmarkDeepSeek V4 ProKimi K2.5Result
HMMT Feb 2026Source 31.7%87.1%Kimi K2.5 leads
IMOAnswerBenchSource 35.3%—Not comparable
ApexSource 0.4%—Not comparable
Apex ShortlistSource 9.2%—Not comparable
AIME 2025Source —96.1%Not comparable
AIME26Source —95.8%Not comparable
AIME25 (Arcee)Source —96.3%Not comparable
HMMT Feb 2025Source —95.4%Not comparable
HMMT Nov 2025Source —91.1%Not comparable
MMAnswerBenchSource —81.8%Not comparable
FrontierMath v2 (Tiers 1-3)Source —27.900%Not comparable
FrontierMath v2 (Tier 4)Source —4.200%Not comparable
Multilingual
BenchmarkDeepSeek V4 ProKimi K2.5Result
MMLU-ProXSource —82.3%Not comparable
NOVA-63Source —56.0%Not comparable
Multimodal
BenchmarkDeepSeek V4 ProKimi K2.5Result
Design Arena WebsiteSource 12641279Kimi K2.5 leads
MMMU-ProSource —78.5%Not comparable
Video-MMESource —87.4%Not comparable
MMVUSource —80.4%Not comparable
VideoMMMUSource —86.6%Not comparable
AA-MMMU-ProSource —75.4%Not comparable
Inst. Following
BenchmarkDeepSeek V4 ProKimi K2.5Result
IFEvalSource —93.9%Not comparable
AA-IFBenchSource —70.2%Not comparable
Frequently Asked Questions (5)

Which is better, DeepSeek V4 Pro or Kimi K2.5?

DeepSeek V4 Pro is ahead on BenchLM's BenchAlign leaderboard, 60.66 to 59.66. The biggest single separator in this matchup is HMMT Feb 2026, where the scores are 31.7% and 87.1%.

Which is better for knowledge tasks, DeepSeek V4 Pro or Kimi K2.5?

Kimi K2.5 has the edge for knowledge tasks in this comparison, averaging 56.9 versus 41.3. Inside this category, HLE is the benchmark that creates the most daylight between them.

Which is better for coding, DeepSeek V4 Pro or Kimi K2.5?

DeepSeek V4 Pro has the edge for coding in this comparison, averaging 65.3 versus 59.4. Inside this category, SWE-bench Verified is the benchmark that creates the most daylight between them.

Which is better for math, DeepSeek V4 Pro or Kimi K2.5?

Kimi K2.5 has the edge for math in this comparison, averaging 60.6 versus 31.7. Inside this category, HMMT Feb 2026 is the benchmark that creates the most daylight between them.

Which is better for agentic tasks, DeepSeek V4 Pro or Kimi K2.5?

DeepSeek V4 Pro has the edge for agentic tasks in this comparison, averaging 59.1 versus 55. Inside this category, MCP Atlas is the benchmark that creates the most daylight between them.

Self-host vs API cost

Estimates at 50,000 req/day · 1000 tokens/req average.

DeepSeek V4 Pro
API / mo$979
Self-host / moNot listed
Break-even—
Proprietary model — self-hosting not applicable.
Kimi K2.5
API / mo$2,700
Self-host / mo$5,221
Break-even132M/day
Model the full break-even

Related Comparisons

Explore More

DeepSeek Compare Pricing Methodology Find Your Best LLM Overall Rankings
Last updated: July 23, 2026

Choose a model with this week’s evidence

Join 2,000+ readers for ranking moves, pricing changes, and the claims that still need proof.

One email each week. Unsubscribe anytime.