← All models
Compare

GPT-5.2 vs Kimi K3

Add, remove, or swap models to compare them side by side.

GPT-5.2Kimi K3
AttributeGPT-5.2OpenAI · 7 in the newsKimi K3Moonshot · 22 in the news
Scores
Intelligence (ECI)154157
Coding6899
Math7983
Reasoning & Knowledge6079
Agentic & Tools7486
Specifications
DeveloperOpenAIMoonshot
FamilyGPTKimi
ReleasedDec 11, 2025Jul 16, 2026
Parameters2.8T
AvailabilityAPI accessOpen weights (non-commercial)
Context window400K1M
Price — $/M input$1.75$3.00
Price — $/M output$14.00$14.00
Inputsfile, image, texttext, image
Outputstexttext
Benchmarks
AIME 2024/202596%97%
APEX34%39%
ARC-AGI86%95%
ARC-AGI-253%60%
FrontierMath41%
FrontierMath Tier 419%
GDPval (win/tie rate)71%
GPQA Diamond91%93%
GSO (code optimization)27%
Humanity's Last Exam28%44%
LiveCodeBench89%
METR task horizon5.9 h
MMLU-Pro87%
SciCode52%59%
SimpleBench46%
SimpleQA Verified39%43%
SWE-bench Verified74%
Terminal-Bench65%
WebDev Arena14801682
WeirdML72%83%
τ²-bench85%

Use-case scores are 0–100 percentile composites across each area’s benchmarks, ranked against every model from the past year. Highlighted cells lead each row. Open a model for the full picture.

Benchmarks & model data from Epoch AI (CC BY); pricing & specs from OpenRouter. ECI = Epoch Capabilities Index.