Skip to main content
BenchLM
Advertise

Model comparison

Kimi K2.6 vs MiniMax M2.7

Data verified

Head-to-head evidence from 26 shared benchmark results across 6 categories. Overall scores shown here use BenchLM's provisional ranking lane.

Kimi K2.6
Moonshot AI
74/100
Margin
19.0pts
← winning
55/100
2 category wins0 category wins

Verified leaderboard positions: Kimi K2.6 #13; MiniMax M2.7 unranked

Evidence parity. Kimi K2.6 and MiniMax M2.7 share 26 comparable benchmark results. 2 of 8 categories are comparable. 34 results are unique to Kimi K2.6; 11 to MiniMax M2.7.

Updated July 13, 2026
Shared results
26
Kimi K2.6 only
34
MiniMax M2.7 only
11
Comparable categories
2 / 8

Pick Kimi K2.6 if you want the stronger benchmark profile. MiniMax M2.7 only becomes the better choice if you want the cheaper token bill or you would rather avoid the extra latency and token burn of a reasoning model.

Confidence note. This is a partial-evidence comparison with 26 shared benchmark results across 6 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 K2.6 is clearly ahead on the provisional aggregate, 74 to 55. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.

Kimi K2.6's sharpest advantage is in coding, where it averages 72.6 against 54.4. The single biggest benchmark swing on the page is Terminal-Bench 2.0, 66.7% to 57%.

Kimi K2.6 is also the more expensive model on tokens at $0.95 input / $4.00 output per 1M tokens, versus $0.30 input / $1.20 output per 1M tokens for MiniMax M2.7. That is roughly 3.3x on output cost alone. Kimi K2.6 is the reasoning model in the pair, while MiniMax M2.7 is not. That usually helps on harder chain-of-thought-heavy tests, but it can also mean more latency and more token spend in real use. Kimi K2.6 gives you the larger context window at 256K, compared with 200K for MiniMax M2.7.

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 Kimi K2.6 and MiniMax M2.7
CategoryKimi K2.6ΔMiniMax M2.7
CodingKimi K2.672.6Margin← 18.2MiniMax M2.754.4
AgenticKimi K2.673.5Margin← 16.5MiniMax M2.757.0
KnowledgeKimi K2.642.2MarginNo overlapMiniMax M2.7Not measured
MathKimi K2.667.1MarginNo overlapMiniMax M2.7Not measured
MultimodalKimi K2.679.8MarginNo overlapMiniMax M2.7Not measured

Decisive benchmark drivers

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

More
A · Kimi K2.6B · MiniMax M2.7

  1. Terminal-Bench 2.0

    Agentic
    Source ↗
    A 66.7%B 57%
    Winner: Kimi K2.6Δ 9.7
    Terminal-Bench 2.0: Kimi K2.6 scored 66.7%; MiniMax M2.7 scored 57%. Kimi K2.6 wins this benchmark.

  2. SWE-bench Pro

    Coding
    Source ↗
    A 58.6%B 56.2%
    Winner: Kimi K2.6Δ 2.4
    SWE-bench Pro: Kimi K2.6 scored 58.6%; MiniMax M2.7 scored 56.2%. Kimi K2.6 wins this benchmark.

Operational comparison

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

MetricKimi K2.6MiniMax M2.7Comparison
Input / output priceUSD per 1M tokensKimi K2.6$0.95 input / $4 outputMiniMax M2.7$0.3 input / $1.2 outputMiniMax M2.7 has the lower combined listed price.
Generation speedtokens per secondKimi K2.6Not availableMiniMax M2.745 tok/sA complete speed comparison is not available.
First-answer latencyseconds to first tokenKimi K2.6Not availableMiniMax M2.72.53 sA complete latency comparison is not available.
Context windowmaximum listed tokensKimi K2.6256KMiniMax M2.7200KKimi K2.6 lists the larger context window.

Benchmark Deep Dive

AgenticKimi K2.6 wins
BenchmarkKimi K2.6MiniMax M2.7Result
Terminal-Bench 2.0Source 66.7%57%Kimi K2.6 leads
BrowseCompSource 83.2%—Not comparable
OSWorld-VerifiedSource 73.1%—Not comparable
ToolathlonSource 50%46.3%Kimi K2.6 leads
MCP AtlasSource 55.9%—Not comparable
Claw-EvalSource 62.3%48.7%Kimi K2.6 leads
DeepSearchQASource 92.5%—Not comparable
WideResearchSource 80.8%—Not comparable
AA Agentic IndexSource 30.3%25.6%Kimi K2.6 leads
Tau2-TelecomSource 95.9%84.8%Kimi K2.6 leads
GDPval-AASource 34.5%33.0%Kimi K2.6 leads
GDPval-AASource 11901160Kimi K2.6 leads
APEX-Agents-AASource 28.5%10.6%Kimi K2.6 leads
Gert LabsSource 56.82%40.40%Kimi K2.6 leads
ResearchClawBenchSource 18.0%—Not comparable
OSWorld 2.0Source 4.6%—Not comparable
AA BriefcaseSource 809—Not comparable
AA AutomationBenchSource 19.6%—Not comparable
AA EnterpriseOps-GymSource 38.5%—Not comparable
AA Harvey LABSource 0.0%—Not comparable
AA ITBenchSource 31.2%—Not comparable
AA Tau3 BankingSource 20.6%—Not comparable
MLE-Bench LiteSource —66.6%Not comparable
MM-ClawBenchSource —62.7%Not comparable
CodingKimi K2.6 wins
BenchmarkKimi K2.6MiniMax M2.7Result
SWE-bench VerifiedSource 80.2%—Not comparable
LiveCodeBenchSource 89.6%—Not comparable
LiveCodeBench v6Source 89.6%—Not comparable
SWE-bench ProSource 58.6%56.2%Kimi K2.6 leads
SWE MultilingualSource 76.7%76.5%Kimi K2.6 leads
SciCodeSource 52.2%—Not comparable
Terminal-Bench 2.0Source 66.7%—Not comparable
Vibe Code BenchSource 37.89%27.04%Kimi K2.6 leads
cursorBench31Source 47.6%—Not comparable
AA Coding IndexSource 61.8%52.6%Kimi K2.6 leads
Terminal-Bench HardSource 43.9%39.4%Kimi K2.6 leads
AA-SciCodeSource 53.5%47.0%Kimi K2.6 leads
AA Terminal-Bench 2.1Source 65.9%—Not comparable
SWE-bench Verified*Source —75.4%Not comparable
SWE-RebenchSource —51.9%Not comparable
Multi-SWE BenchSource —52.7%Not comparable
VIBE-ProSource —55.6%Not comparable
NL2RepoSource —39.8%Not comparable
React Native EvalsSource —71.4%Not comparable
Reasoning
BenchmarkKimi K2.6MiniMax M2.7Result
AA-LCRSource 69.7%68.7%Kimi K2.6 leads
CritPtSource 8.0%0.6%Kimi K2.6 leads
Knowledge
BenchmarkKimi K2.6MiniMax M2.7Result
GPQASource 90.5%—Not comparable
GPQA-DSource 90.5%87.0%Kimi K2.6 leads
HLESource 34.7%—Not comparable
Artificial Analysis Intelligence IndexSource 44.2%38.1%Kimi K2.6 leads
AA-GPQA DiamondSource 91.1%87.4%Kimi K2.6 leads
AA-HLESource 35.9%28.1%Kimi K2.6 leads
AA-Omniscience IndexSource 6.4%0.7%Kimi K2.6 leads
AA-Omniscience AccuracySource 32.8%26.1%Kimi K2.6 leads
AA-Omniscience Hallucination RateSource 39.3%34.4%MiniMax M2.7 leads
AA Openness IndexSource 33.3%—Not comparable
MMLU-Pro (Arcee)Source —80.8%Not comparable
Math
BenchmarkKimi K2.6MiniMax M2.7Result
AIME26Source 96.4%—Not comparable
HMMT Feb 2026Source 92.7%—Not comparable
MMAnswerBenchSource 86.0%—Not comparable
FrontierMath v2 (Tiers 1-3)Source 38.966%—Not comparable
FrontierMath v2 (Tier 4)Source 14.580%—Not comparable
AIME25 (Arcee)Source —80.0%Not comparable
Multimodal
BenchmarkKimi K2.6MiniMax M2.7Result
MMMU-ProSource 79.4%—Not comparable
MMMU-Pro w/ PythonSource 80.1%—Not comparable
CharXivSource 80.4%—Not comparable
MathVisionSource 87.4%—Not comparable
V*Source 96.9%—Not comparable
AA-MMMU-ProSource 79.4%—Not comparable
Design Arena WebsiteSource 13181287Kimi K2.6 leads
GDPval-AASource —1495Not comparable
Inst. Following
BenchmarkKimi K2.6MiniMax M2.7Result
AA-IFBenchSource 76.0%75.7%Kimi K2.6 leads
Frequently Asked Questions (3)

Which is better, Kimi K2.6 or MiniMax M2.7?

Kimi K2.6 is ahead on BenchLM's provisional leaderboard, 74 to 55. The biggest single separator in this matchup is Terminal-Bench 2.0, where the scores are 66.7% and 57%.

Which is better for coding, Kimi K2.6 or MiniMax M2.7?

Kimi K2.6 has the edge for coding in this comparison, averaging 72.6 versus 54.4. Inside this category, Vibe Code Bench is the benchmark that creates the most daylight between them.

Which is better for agentic tasks, Kimi K2.6 or MiniMax M2.7?

Kimi K2.6 has the edge for agentic tasks in this comparison, averaging 73.5 versus 57. Inside this category, GDPval-AA 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.

Kimi K2.6
API / mo$3,713
Self-host / mo$18,221
Break-even326M/day
MiniMax M2.7
API / mo$1,125
Self-host / moNot listed
Break-even—
Proprietary model — self-hosting not applicable.
Model the full break-even

Related Comparisons

Explore More

Moonshot AI Compare Pricing Methodology Find Your Best LLM Overall Rankings
Last updated: July 13, 2026

The AI models change fast. We track them for you.

A weekly brief for engineers and researchers covering new models, ranking shifts, and pricing changes.

Free. No spam. Unsubscribe anytime.