Skip to main content
BenchLM
Advertise

Model comparison

Kimi K2.5 vs MAI-Thinking-1

Data verified

Head-to-head evidence from 9 shared benchmark results across 4 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.

Kimi K2.5
Moonshot AI
59.66/100
No comparison
N/A
2 category wins3 category wins

Public leaderboard positions: Kimi K2.5 #54 (Supported); MAI-Thinking-1 unranked (Not scored). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.

Evidence parity. Kimi K2.5 and MAI-Thinking-1 share 9 comparable benchmark results. 5 of 8 categories are comparable. 54 results are unique to Kimi K2.5; 4 to MAI-Thinking-1.

Updated July 23, 2026
Shared results
9
Kimi K2.5 only
54
MAI-Thinking-1 only
4
Comparable categories
5 / 8

Treat this as a split decision. Kimi K2.5 makes more sense if agentic is the priority or you would rather avoid the extra latency and token burn of a reasoning model; MAI-Thinking-1 is the better fit if mathematics is the priority or you want the stronger reasoning-first profile.

Confidence note. This is a partial-evidence comparison with 9 shared benchmark results across 4 evidence categories; 5 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.

Why this result

Kimi K2.5 and MAI-Thinking-1 finish on the same BenchAlign overall score, so this is less about a single winner and more about where the edge shows up. The BenchAlign headline says tie; the benchmark table is where the real choice happens.

MAI-Thinking-1 is the reasoning model in the pair, while Kimi K2.5 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.

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 K2.5 and MAI-Thinking-1
CategoryKimi K2.5ΔMAI-Thinking-1
MathKimi K2.560.6Margin→ 29.1MAI-Thinking-189.7
KnowledgeKimi K2.556.9Margin→ 15.6MAI-Thinking-172.5
AgenticKimi K2.555.0Margin← 9.0MAI-Thinking-146.0
Inst. FollowingKimi K2.593.9Margin← 8.9MAI-Thinking-185.0
CodingKimi K2.559.4Margin→ 6.1MAI-Thinking-165.5
ReasoningKimi K2.561.0MarginNo overlapMAI-Thinking-1Not measured
MultilingualKimi K2.582.3MarginNo overlapMAI-Thinking-1Not measured
MultimodalKimi K2.578.5MarginNo overlapMAI-Thinking-1Not measured

Decisive benchmark drivers

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

More
A · Kimi K2.5B · MAI-Thinking-1

  1. Terminal-Bench 2.0

    Agentic
    Source ↗
    A 50.8%B 46%
    Winner: Kimi K2.5Δ 4.8
    Terminal-Bench 2.0: Kimi K2.5 scored 50.8%; MAI-Thinking-1 scored 46%. Kimi K2.5 wins this benchmark.

  2. GPQA

    Knowledge
    Source ↗
    A 87.6%B 84.2%
    Winner: Kimi K2.5Δ 3.4
    GPQA: Kimi K2.5 scored 87.6%; MAI-Thinking-1 scored 84.2%. Kimi K2.5 wins this benchmark.

  3. SWE-bench Verified

    Coding
    Source ↗
    A 76.8%B 73.5%
    Winner: Kimi K2.5Δ 3.3
    SWE-bench Verified: Kimi K2.5 scored 76.8%; MAI-Thinking-1 scored 73.5%. Kimi K2.5 wins this benchmark.

  4. HMMT Feb 2026

    Math
    Source ↗
    A 87.1%B 84.9%
    Winner: Kimi K2.5Δ 2.2
    HMMT Feb 2026: Kimi K2.5 scored 87.1%; MAI-Thinking-1 scored 84.9%. Kimi K2.5 wins this benchmark.

  5. SWE-bench Pro

    Coding
    Source ↗
    A 50.7%B 52.8%
    Winner: MAI-Thinking-1Δ 2.1
    SWE-bench Pro: Kimi K2.5 scored 50.7%; MAI-Thinking-1 scored 52.8%. MAI-Thinking-1 wins this benchmark.

Operational comparison

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

MetricKimi K2.5MAI-Thinking-1Comparison
Input / output priceUSD per 1M tokensKimi K2.5$0.6 input / $3 outputMAI-Thinking-1Not availableA complete price comparison is not available.
Generation speedtokens per secondKimi K2.545 tok/sMAI-Thinking-1Not availableA complete speed comparison is not available.
First-answer latencyseconds to first tokenKimi K2.52.38 sMAI-Thinking-1Not availableA complete latency comparison is not available.
Context windowmaximum listed tokensKimi K2.5256KMAI-Thinking-1256KListed context windows are equal.

Benchmark Deep Dive

AgenticKimi K2.5 wins
BenchmarkKimi K2.5MAI-Thinking-1Result
Terminal-Bench 2.0Source 50.8%46%Kimi K2.5 leads
BrowseCompSource 60.6%—Not comparable
Claw-EvalSource 52.3%—Not comparable
QwenClawBenchSource 54.3%—Not comparable
τ³-bench resultsSource 65.7%—Not comparable
DeepSearchQASource 77.1%—Not comparable
DeepPlanningSource 14.4%—Not comparable
ToolathlonSource 27.8%—Not comparable
MCP AtlasSource 29.5%—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
Gert LabsSource 45.88%—Not comparable
ResearchClawBenchSource 14.0%—Not comparable
JobBenchSource 8.7%—Not comparable
AA Agentic IndexSource 21.7%—Not comparable
GDPval-AASource 25.4%—Not comparable
GDPval-AASource 1009—Not comparable
CodingMAI-Thinking-1 wins
BenchmarkKimi K2.5MAI-Thinking-1Result
SWE-bench VerifiedSource 76.8%73.5%Kimi K2.5 leads
SWE-bench Verified*Source 70.8%—Not comparable
LiveCodeBench v6Source 85.0%—Not comparable
SWE-bench ProSource 50.7%52.8%MAI-Thinking-1 leads
SWE MultilingualSource 73%—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
Terminal-Bench 2.0Source —46.0%Not comparable
Reasoning
BenchmarkKimi K2.5MAI-Thinking-1Result
LongBench v2Source 61%—Not comparable
AA-LCRSource 65.3%—Not comparable
CritPtSource 3.1%—Not comparable
Graphwalks BFS 128KSource —90%Not comparable
KnowledgeMAI-Thinking-1 wins
BenchmarkKimi K2.5MAI-Thinking-1Result
GPQASource 87.6%84.2%Kimi K2.5 leads
GPQA-DSource 87.6%84.2%Kimi K2.5 leads
SuperGPQASource 69.2%—Not comparable
MMLU-ProSource 87.1%85%Kimi K2.5 leads
MMLU-Pro (Arcee)Source 87.1%—Not comparable
HLESource 30.1%—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
SimpleQASource —31%Not comparable
MathMAI-Thinking-1 wins
BenchmarkKimi K2.5MAI-Thinking-1Result
AIME 2025Source 96.1%97%MAI-Thinking-1 leads
AIME26Source 95.8%94.5%Kimi K2.5 leads
AIME25 (Arcee)Source 96.3%—Not comparable
HMMT Feb 2025Source 95.4%—Not comparable
HMMT Nov 2025Source 91.1%—Not comparable
HMMT Feb 2026Source 87.1%84.9%Kimi K2.5 leads
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
BenchmarkKimi K2.5MAI-Thinking-1Result
MMLU-ProXSource 82.3%—Not comparable
NOVA-63Source 56.0%—Not comparable
Multimodal
BenchmarkKimi K2.5MAI-Thinking-1Result
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
Design Arena WebsiteSource 1279—Not comparable
Inst. FollowingKimi K2.5 wins
BenchmarkKimi K2.5MAI-Thinking-1Result
IFEvalSource 93.9%—Not comparable
AA-IFBenchSource 70.2%—Not comparable
IFBenchSource —85%Not comparable
Frequently Asked Questions (6)

Which is better, Kimi K2.5 or MAI-Thinking-1?

Kimi K2.5 and MAI-Thinking-1 are tied on the BenchAlign overall score, so the right pick depends on which category matters most for your use case.

Which is better for knowledge tasks, Kimi K2.5 or MAI-Thinking-1?

MAI-Thinking-1 has the edge for knowledge tasks in this comparison, averaging 72.5 versus 56.9. Inside this category, GPQA is the benchmark that creates the most daylight between them.

Which is better for coding, Kimi K2.5 or MAI-Thinking-1?

MAI-Thinking-1 has the edge for coding in this comparison, averaging 65.5 versus 59.4. Inside this category, SWE-bench Verified is the benchmark that creates the most daylight between them.

Which is better for math, Kimi K2.5 or MAI-Thinking-1?

MAI-Thinking-1 has the edge for math in this comparison, averaging 89.7 versus 60.6. Inside this category, HMMT Feb 2026 is the benchmark that creates the most daylight between them.

Which is better for agentic tasks, Kimi K2.5 or MAI-Thinking-1?

Kimi K2.5 has the edge for agentic tasks in this comparison, averaging 55 versus 46. Inside this category, Terminal-Bench 2.0 is the benchmark that creates the most daylight between them.

Which is better for instruction following, Kimi K2.5 or MAI-Thinking-1?

Kimi K2.5 has the edge for instruction following in this comparison, averaging 93.9 versus 85. MAI-Thinking-1 stays close enough that the answer can still flip depending on your workload.

Self-host vs API cost

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

Kimi K2.5
API / mo$2,700
Self-host / mo$5,221
Break-even132M/day
MAI-Thinking-1
API / mo$0
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 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.