DataLearner logo

Kimi K3vsMiniMax M3

Across 3 shared benchmarks, Kimi K3 leads overall: Kimi K3 wins 3, MiniMax M3 wins 0, with 0 ties and an average score difference of +14.93.

Moonshot AI
Kimi K3

Moonshot AI · 2026-07-16 · Reasoning model

MiniMaxAI
MiniMax M3

MiniMaxAI · 2026-06-01 · Multimodal model

Kimi K33 wins(100%)(0%)0 winsMiniMax M3

Benchmark scores

Grouped by capability, sorted by largest gap within each. 3 shared benchmarks.

AI Agent - Tool Usage

Kimi K3 2/2
BenchmarkKimi K3MiniMax M3Diff
TerminalBench 2.188.302 / 27Max (With Tools)6621 / 27Thinking (With Tools)+22.30
OSWorld-Verified84.802 / 24Max (With Tools)7018 / 24Thinking (With Tools)+14.80

AI Agent - Information Search

Kimi K3 1/1
BenchmarkKimi K3MiniMax M3Diff
BrowseComp91.201 / 53Max (With Tools + Internet)83.5012 / 53Thinking (With Tools + Internet)+7.70

Specs

FieldKimi K3MiniMax M3
PublisherMoonshot AIMiniMaxAI
Release date2026-07-162026-06-01
Model typeReasoning modelMultimodal model
ArchitectureMoEMoE
Parameters2.8T428B
Context length1M1M
Max output1M512K

API pricing

Prices use DataLearner records when available; missing fields are not inferred.

ItemKimi K3MiniMax M3
Text input¥20 / 1M tokens¥2.1 / 1M tokens
Text output¥100 / 1M tokens¥8.4 / 1M tokens
Cache read¥2 / 1M tokens¥0.42 / 1M tokens

Summary

  • Kimi K3leads in:AI Agent - Tool Usage (2/2), AI Agent - Information Search (1/1)

On average across the 3 shared benchmarks, Kimi K3 scores 14.93 higher.

Largest single-benchmark gap: TerminalBench 2.1 — Kimi K3 88.30 vs MiniMax M3 66 (+22.30).

Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.