Skip to main content
BenchLM
Advertise

Model comparison

GLM-5.2 vs Kimi K3

Data verified

Head-to-head evidence from 29 shared benchmark results across 5 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.

63.96/100
Margin
17.0pts
winning →
Kimi K3
Moonshot AI
80.96/100
0 category wins2 category wins

Public leaderboard positions: GLM-5.2 #37 (Estimated); Kimi K3 #4 (Supported). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.

Evidence parity. GLM-5.2 and Kimi K3 share 29 comparable benchmark results. 2 of 8 categories are comparable. 14 results are unique to GLM-5.2; 28 to Kimi K3.

Updated July 23, 2026
Shared results
29
GLM-5.2 only
14
Kimi K3 only
28
Comparable categories
2 / 8

Pick Kimi K3 if you want the stronger benchmark profile. GLM-5.2 only becomes the better choice if you want the cheaper token bill.

Confidence note. This is a partial-evidence comparison with 29 shared benchmark results across 5 evidence categories; 2 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.

Why this result

Kimi K3 is clearly ahead on the BenchAlign aggregate, 80.96 to 63.96. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.

Kimi K3's sharpest advantage is in agentic, where it averages 89.5 against 81. The single biggest benchmark swing on the page is Terminal-Bench 2.0, 81% to 88.3%.

Kimi K3 is also the more expensive model on tokens at $3.00 input / $15.00 output per 1M tokens, versus $1.40 input / $4.40 output per 1M tokens for GLM-5.2. That is roughly 3.4x on output cost alone. Kimi K3 gives you the larger context window at 1.05M, compared with 1M for GLM-5.2.

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 GLM-5.2 and Kimi K3
CategoryGLM-5.2ΔKimi K3
AgenticGLM-5.281.0Margin→ 8.5Kimi K389.5
KnowledgeGLM-5.259.6Margin→ 1.4Kimi K361.0
CodingGLM-5.262.1MarginNo overlapKimi K3Not measured
MathGLM-5.295.9MarginNo overlapKimi K3Not measured
MultimodalGLM-5.2Not measuredMarginNo overlapKimi K378.5

Decisive benchmark drivers

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

More
A · GLM-5.2B · Kimi K3

  1. Terminal-Bench 2.0

    Agentic
    Source ↗
    A 81%B 88.3%
    Winner: Kimi K3Δ 7.3
    Terminal-Bench 2.0: GLM-5.2 scored 81%; Kimi K3 scored 88.3%. Kimi K3 wins this benchmark.

  2. GPQA

    Knowledge
    Source ↗
    A 91.2%B 93.5%
    Winner: Kimi K3Δ 2.3
    GPQA: GLM-5.2 scored 91.2%; Kimi K3 scored 93.5%. Kimi K3 wins this benchmark.

  3. HLE

    Knowledge
    Source ↗
    A 54.7%B 56%
    Winner: Kimi K3Δ 1.3
    HLE: GLM-5.2 scored 54.7%; Kimi K3 scored 56%. Kimi K3 wins this benchmark.

Operational comparison

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

MetricGLM-5.2Kimi K3Comparison
Input / output priceUSD per 1M tokensGLM-5.2$1.4 input / $4.4 outputKimi K3$3 input / $15 outputGLM-5.2 has the lower combined listed price.
Generation speedtokens per secondGLM-5.2Not availableKimi K3Not availableA complete speed comparison is not available.
First-answer latencyseconds to first tokenGLM-5.2Not availableKimi K3Not availableA complete latency comparison is not available.
Context windowmaximum listed tokensGLM-5.21MKimi K31.05MKimi K3 lists the larger context window.

Benchmark Deep Dive

AgenticKimi K3 wins
BenchmarkGLM-5.2Kimi K3Result
Terminal-Bench 2.0Source 81%88.3%Kimi K3 leads
MCP AtlasSource 76.8%84.2%Kimi K3 leads
ToolathlonSource 48.2%—Not comparable
AA Agentic IndexSource 43.1%50.1%Kimi K3 leads
τ²-bench resultsSource 99.1%—Not comparable
GDPval-AASource 50.7%59.0%Kimi K3 leads
GDPval-AASource 15141679Kimi K3 leads
APEX-Agents-AASource 33.7%41.3%Kimi K3 leads
ResearchClawBenchSource 20.7%—Not comparable
AA BriefcaseSource 12601543Kimi K3 leads
AA AutomationBenchSource 27.8%52.7%Kimi K3 leads
AA EnterpriseOps-GymSource 42.7%45.3%Kimi K3 leads
AA Harvey LABSource 91.0%94.6%Kimi K3 leads
AA ITBenchSource 42.7%47.7%Kimi K3 leads
AA Tau3 BankingSource 26.8%33.4%Kimi K3 leads
terminalBenchHardSource 50.8%—Not comparable
aaTerminalBench21Source 77.9%85%Kimi K3 leads
BrowseCompSource —91.2%Not comparable
DeepSearchQASource —95.0%Not comparable
Toolathlon-VerifiedSource —73.2%Not comparable
AutomationBenchSource —30.8%Not comparable
JobBenchSource —52.9%Not comparable
APEX-AgentsSource —37.6%Not comparable
SpreadsheetBench 2Source —34.8%Not comparable
DECK-BenchSource —73.5%Not comparable
Coding
BenchmarkGLM-5.2Kimi K3Result
SWE-bench ProSource 62.1%—Not comparable
NL2RepoSource 48.9%—Not comparable
Terminal-Bench 2.0Source 81.0%—Not comparable
ProgramBenchSource 63.7%77.8%Kimi K3 leads
cursorBench32Source 55.0%—Not comparable
AA Coding IndexSource 68.8%76.2%Kimi K3 leads
AA-SciCodeSource 50.5%58.7%Kimi K3 leads
deepSweSource —67.5%Not comparable
FrontierSWESource —81.2%Not comparable
Kimi Code Bench v2Source —72.9%Not comparable
sweMarathonSource —42%Not comparable
PostTrain BenchSource —36.6%Not comparable
MLS-Bench LiteSource —48.3%Not comparable
Reasoning
BenchmarkGLM-5.2Kimi K3Result
CritPtSource 20.9%23.4%Kimi K3 leads
AA-LCRSource 71.3%74.7%Kimi K3 leads
KnowledgeKimi K3 wins
BenchmarkGLM-5.2Kimi K3Result
GPQASource 91.2%93.5%Kimi K3 leads
GPQA-DSource 91.2%93.5%Kimi K3 leads
HLESource 54.7%56%Kimi K3 leads
HLE w/o toolsSource 40.5%43.5%Kimi K3 leads
Artificial Analysis Intelligence IndexSource 51.1%57.1%Kimi K3 leads
AA-GPQA DiamondSource 89.5%93.5%Kimi K3 leads
AA-HLESource 40.1%44.3%Kimi K3 leads
AA-Omniscience IndexSource 4.0%18.4%Kimi K3 leads
AA-Omniscience AccuracySource 25.1%46.0%Kimi K3 leads
AA-Omniscience Hallucination RateSource 28.1%50.9%GLM-5.2 leads
AA Openness IndexSource 44.4%—Not comparable
Math
BenchmarkGLM-5.2Kimi K3Result
AIME26Source 99.2%—Not comparable
HMMT Nov 2025Source 94.4%—Not comparable
HMMT Feb 2026Source 92.5%—Not comparable
MMAnswerBenchSource 91.0%—Not comparable
Multimodal
BenchmarkGLM-5.2Kimi K3Result
Design Arena WebsiteSource 13401386Kimi K3 leads
OfficeQA ProSource —63.3%Not comparable
MMMU-ProSource —81.6%Not comparable
MMMU-Pro w/ PythonSource —83.4%Not comparable
CharXiv w/o toolsSource —84.8%Not comparable
CharXivSource —91.3%Not comparable
MathVisionSource —94.3%Not comparable
MathVision w/ PythonSource —97.8%Not comparable
BabyVision w/ PythonSource —85.7%Not comparable
ZeroBenchSource —23.0%Not comparable
ZeroBench w/ PythonSource —41.0%Not comparable
WorldVQA ForceAnswerSource —51.0%Not comparable
OmniDocBenchSource —91.1%Not comparable
PerceptionBenchSource —58.5%Not comparable
AA-MMMU-ProSource —80.5%Not comparable
Inst. Following
BenchmarkGLM-5.2Kimi K3Result
AA-IFBenchSource 73.3%—Not comparable
Frequently Asked Questions (3)

Which is better, GLM-5.2 or Kimi K3?

Kimi K3 is ahead on BenchLM's BenchAlign leaderboard, 80.96 to 63.96. The biggest single separator in this matchup is Terminal-Bench 2.0, where the scores are 81% and 88.3%.

Which is better for knowledge tasks, GLM-5.2 or Kimi K3?

Kimi K3 has the edge for knowledge tasks in this comparison, averaging 61 versus 59.6. Inside this category, AA-Omniscience Hallucination Rate is the benchmark that creates the most daylight between them.

Which is better for agentic tasks, GLM-5.2 or Kimi K3?

Kimi K3 has the edge for agentic tasks in this comparison, averaging 89.5 versus 81. Inside this category, AA Briefcase is the benchmark that creates the most daylight between them.

Related Comparisons

Explore More

Z.AI 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.