Kimi K3 vs MiniMax M3

Updated October 2, 2026. Rank says Kimi K3 is ahead. Price, access, and your workload can each overturn that. Public scores include evidence status and uncertainty.

Decision reading

Kimi K3 has the higher public score, 72.14 versus 55.18, and the 90% score intervals do not overlap. 11 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.

Model A

Kimi K3

Moonshot AI

72.14/100

Supported · Public rank #15

90% interval 68.8–75.5

Model B

MiniMax M3

MiniMax

55.18/100

Supported · Public rank #66

90% interval 46.3–64.0

Shared results: 11
Kimi K3 only: 37
MiniMax M3 only: 16
Like-for-like categories: 3 / 8

Supported: Kimi K3 and MiniMax M3How the comparison works

Which one for your work

Recommendations appear only when a shared evidence basis or an explicit operating constraint supports the call. Secondary and unsupported use cases stay disclosed below the initial list.

Coding work
Code generation, repair, and software-engineering tasks
Kimi K3
Kimi K3 has the higher public coding point estimate, 61.4 to 38.7, with Supported evidence for both models. The conditional ranges do not establish rank confidence.
Confidence: limited
Agentic work
Tool use, computer use, and multi-step task completion
Kimi K3
Kimi K3 has the higher public agentic point estimate, 68.1 to 39.8, with Supported evidence for both models. The conditional ranges do not establish rank confidence.
Confidence: limited
Long documents
Prompts that approach the documented context limit
Kimi K3
Kimi K3 has the larger documented context window.
Confidence: documented

Show secondary and unsupported calls

Chat turn cost
1K fresh input + 500 output tokens
MiniMax M3
MiniMax M3 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
Cache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
MiniMax M3
MiniMax M3 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
Repository review cost
50K fresh input + 3K output tokens
MiniMax M3
MiniMax M3 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates

Which one for a specific job

Choose a job from the LLM Selector's task catalog to see the category row it rests on, under the same basis rules as the table below. A directional row stays directional; choosing a job never creates a winner.

Pick the job you would use them for

The same task catalog as the LLM Selector. Each job names the evidence surface it rests on; nothing here adds a new score.

61.4Kimi K338.7MiniMax M3

Like-for-like · BenchAlign v5.8

Kimi K3 has the higher coding point estimate. Conditional score ranges do not establish rank confidence.

Coding scores combine specific tasks and setups. Match the editor, harness, and effort to your workflow.

Open the coding leaderboard Shortlist models for this job

Same basis rules as the category table below

Shape of the matched evidence

Only shared public evidence is shown. Sparse evidence stays a ruled list rather than being closed into a radar shape.

Too few matched category axes support a radar. The ruled list below shows only shared benchmark results; positions use each benchmark’s normalized display scale when available.

OfficeQA ProMultimodal
Normalized gap 18.2
Kimi K3:63.3%
MiniMax M3:45.1%
Kimi K3 source MiniMax M3 source
BrowseCompAgentic
Normalized gap 7.7
Kimi K3:91.2%
MiniMax M3:83.5%
Kimi K3 source MiniMax M3 source
LiveCodeBench (Vals)Coding
Normalized gap 5.0
Kimi K3:87.2%
MiniMax M3:82.2%
Kimi K3 source MiniMax M3 source
MMLU-Pro (Vals)Knowledge
Normalized gap 3.8
Kimi K3:88.0%
MiniMax M3:84.2%
Kimi K3 source MiniMax M3 source
MMMU-ProMultimodal
Normalized gap 3.5
Kimi K3:81.6%
MiniMax M3:78.1%
Kimi K3 source MiniMax M3 source

Bars run 0–100 on each benchmark’s normalized display scale

Category results, on a stated basis

Each row shows the public-lane category score for both models: the BenchAlign v5.8 lane for agentic, coding, and knowledge, and the weighted public rows elsewhere, with the same rank each model page reports. A row is like-for-like only when both scores rest on Supported evidence or the same weighted set. Directional and not-comparable rows remain visible, but they never receive a winner.

Agentic

Like-for-like

Kimi K3: 68.1; Supported · #8/119
MiniMax M3: 39.8; Supported · #58/119
Basis: BenchAlign v5.8 lane · 12 vs 9 public rows
Reading: Kimi K3 leads

Coding

Like-for-like

Kimi K3: 61.4; Supported · #18/144
MiniMax M3: 38.7; Supported · #69/144
Basis: BenchAlign v5.8 lane · 14 vs 10 public rows
Reading: Kimi K3 leads

Knowledge

Like-for-like

Kimi K3: 67.9; Supported · #18/171
MiniMax M3: 48.4; Supported · #73/171
Basis: BenchAlign v5.8 lane · 6 vs 2 public rows
Reading: Kimi K3 leads

Reasoning

Directional only

Kimi K3: 65.8; #18/27
MiniMax M3: 79.4; #5/27
Basis: Provisional lane · 1 vs 0 weighted rows
Reading: Directional only

Multimodal

Directional only

Kimi K3: 89.4; #1/49
MiniMax M3: 51.6; #36/49
Basis: Provisional lane · 3 vs 2 weighted rows
Reading: Directional only

Multilingual

Not comparable

Kimi K3: Not ranked
MiniMax M3: Not ranked
Basis: Provisional lane · 0 vs 0 weighted rows
Reading: Not comparable

Instruction following

Not comparable

Kimi K3: Not ranked
MiniMax M3: 92.4; #5/125
Basis: Provisional lane · 0 vs 0 weighted rows
Reading: Not comparable

Math

Not comparable

Kimi K3: Not ranked
MiniMax M3: Not ranked
Basis: Provisional lane · 0 vs 1 weighted rows
Reading: Not comparable

Category scores with the server-provided evidence basis for Kimi K3 and MiniMax M3
Category	Kimi K3	MiniMax M3	Basis	Reading
Agentic	68.1Supported · #8/119	39.8Supported · #58/119	Like-for-likeBenchAlign v5.8 lane · 12 vs 9 public rows	Kimi K3 leads
Coding	61.4Supported · #18/144	38.7Supported · #69/144	Like-for-likeBenchAlign v5.8 lane · 14 vs 10 public rows	Kimi K3 leads
Knowledge	67.9Supported · #18/171	48.4Supported · #73/171	Like-for-likeBenchAlign v5.8 lane · 6 vs 2 public rows	Kimi K3 leads
Reasoning	65.8#18/27	79.4#5/27	Directional onlyProvisional lane · 1 vs 0 weighted rows	Directional only
Multimodal	89.4#1/49	51.6#36/49	Directional onlyProvisional lane · 3 vs 2 weighted rows	Directional only
Multilingual	Not ranked	Not ranked	Not comparableProvisional lane · 0 vs 0 weighted rows	Not comparable
Instruction following	Not ranked	92.4#5/125	Not comparableProvisional lane · 0 vs 0 weighted rows	Not comparable
Math	Not ranked	Not ranked	Not comparableProvisional lane · 0 vs 1 weighted rows	Not comparable

Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign v5.8) differ from the provisional-lane categories. Unranked scores sit on the provisional lane’s scale but never name a winner. Like-for-like rows within 0.5 points read as a practical tie.

Supported evidence per lane · bars run 0–100Methodology

What each workload costs

Three fixed token mixes turn per-token rates into comparable decisions. Each scenario states context fit and whether cached input had to fall back to the published list-input rate.

Chat turn

1K fresh input + 500 output tokens

Kimi K3: $0.0105; Fits in one request
MiniMax M3: $0.0009; Fits in one request

MiniMax M3 has the lower modeled cost

Costs use the listed standard API rates.

Repository review

50K fresh input + 3K output tokens

Kimi K3: $0.195; Fits in one request
MiniMax M3: $0.0186; Fits in one request

MiniMax M3 has the lower modeled cost

Costs use the listed standard API rates.

Cache-heavy agent loop

200K cached + 20K fresh input + 10K output tokens

Kimi K3: $0.27; Fits in one request
MiniMax M3: $0.03; Fits in one request

MiniMax M3 has the lower modeled cost

Costs use the listed standard API rates.

Cached input falls back to the list input rate only where a cached rate is unpublished

Specification differences

Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.

SpecificationKimi K3MiniMax M3

Context window

Maximum documented context; output-token limits may be lower.

Kimi K3

1.05M

MiniMax M3

API model ID

Kimi K3

Not sourced

MiniMax M3

Not sourced

Cached-input rate

A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.

Kimi K3

$0.3 per 1M cached input tokens

MiniMax M3

$0.06 per 1M cached input tokens

Documented inputs

Kimi K3

Not sourced

MiniMax M3

Not sourced

Documented outputs

Kimi K3

Not sourced

MiniMax M3

Not sourced

Provider availability

Kimi K3

Not sourced

MiniMax M3

Not sourced

Reasoning profile

Kimi K3

Reasoning

MiniMax M3

Non-Reasoning

Weight access

Kimi K3

Pending

MiniMax M3

Open Weight

License

Kimi K3

Pending

MiniMax M3

Open Weight

Release date

Kimi K3

2026-07-16

MiniMax M3

2026-06-01

If you already use one of these models

Deployment change: The models list different providers, so authentication, endpoint behavior, limits, and feature support may change.
Quality signal: Kimi K3 has the higher public score, 72.14 versus 55.18, and the 90% score intervals do not overlap.
Workload cost: Repository review: $0.195 vs $0.0186. Cache-heavy agent loop: $0.27 vs $0.03.
Context tradeoff: Kimi K3 has the larger documented window (1.05M).

Run the same representative tasks against both endpoints before changing production traffic.

Questions

Which is better, Kimi K3 or MiniMax M3?

Kimi K3 has the higher public score, 72.14 versus 55.18, and the 90% score intervals do not overlap. That is the clearest overall quality signal in the current public data.

Which is better for coding, Kimi K3 or MiniMax M3?

Kimi K3 has the higher public coding point estimate, 61.4 to 38.7, with Supported evidence for both models. The conditional ranges do not establish rank confidence.

Which is better for agentic tasks, Kimi K3 or MiniMax M3?

Kimi K3 has the higher public agentic tasks point estimate, 68.1 to 39.8, with Supported evidence for both models. The conditional ranges do not establish rank confidence.

Which costs less, Kimi K3 or MiniMax M3?

For the stated presets, chat costs $0.0105 on Kimi K3 and $0.0009 on MiniMax M3; repository review costs $0.195 and $0.0186; the cache-heavy agent loop costs $0.27 and $0.03. Costs use the listed standard API rates.

Which has the larger context window, Kimi K3 or MiniMax M3?

Kimi K3 has the larger documented context window: 1.05M, compared with 1M.

Benchmark evidence

The full public result ledger is available for audit without forcing a wide desktop table onto a phone.

Browse raw public benchmark evidence64 rows

Agentic

Terminal-Bench 2.1
Kimi K388.3%
Source
MiniMax M366.0%
Source
Kimi K3 leads this result
BrowseComp
Kimi K391.2%
Source
MiniMax M383.5%
Source
Kimi K3 leads this result
DeepSearchQA
Kimi K395.0%
Source
MiniMax M3—
Not directly comparable
Toolathlon-Verified
Kimi K373.2%
Source
MiniMax M3—
Not directly comparable
MCP Atlas
Kimi K384.2%
Source
MiniMax M374.2%
Source
Kimi K3 leads this result
AutomationBench
Kimi K330.8%
Source
MiniMax M3—
Not directly comparable
JobBench
Kimi K352.9%
Source
MiniMax M3—
Not directly comparable
APEX-Agents
Kimi K337.6%
Source
MiniMax M3—
Not directly comparable
SpreadsheetBench 2
Kimi K334.8%
Source
MiniMax M3—
Not directly comparable
DECK-Bench
Kimi K373.5%
Source
MiniMax M3—
Not directly comparable
Terminal-Bench 2.1 (Vals)
Kimi K380.9%
Source
MiniMax M353.6%
Source
Kimi K3 leads this result
ApprenticeBench
Kimi K318%
Source
MiniMax M3—
Not directly comparable
OSWorld-Verified
Kimi K3—
MiniMax M370.1%
Source
Not directly comparable
Claw-Eval
Kimi K3—
MiniMax M374.5%
Source
Not directly comparable
BankerToolBench
Kimi K3—
MiniMax M376.1%
Source
Not directly comparable
ResearchClawBench
Kimi K3—
MiniMax M319.8%
Source
Not directly comparable
OSWorld 2.0
Kimi K3—
MiniMax M34.6%
Source
Not directly comparable

Coding

DeepSWE
Kimi K367.5%
Source
MiniMax M3—
Not directly comparable
CursorBench 3.2
Kimi K360.8%
Source
MiniMax M3—
Not directly comparable
FrontierSWE
Kimi K381.2%
Source
MiniMax M3—
Not directly comparable
ProgramBench
Kimi K377.8%
Source
MiniMax M3—
Not directly comparable
Kimi Code Bench v2
Kimi K372.9%
Source
MiniMax M3—
Not directly comparable
sweMarathon
Kimi K342%
Source
MiniMax M3—
Not directly comparable
PostTrain Bench
Kimi K336.6%
Source
MiniMax M3—
Not directly comparable
MLS-Bench Lite
Kimi K348.3%
Source
MiniMax M3—
Not directly comparable
VulcanBench v3
Kimi K373.7%
Source
MiniMax M3—
Not directly comparable
OpenHarmony Bench
Shared source
Kimi K357.3%
MiniMax M348.4%
Kimi K3 leads this result
FrontierSWE v2
Kimi K325.9%
Source
MiniMax M3—
Not directly comparable
LiveCodeBench (Vals)
Kimi K387.2%
Source
MiniMax M382.2%
Source
Kimi K3 leads this result
SWE-bench (Vals)
Kimi K393.4%
Source
MiniMax M375.0%
Source
Kimi K3 leads this result
PostTrainBench v1.1
Kimi K332.0%
Source
MiniMax M3—
Not directly comparable
SWE-bench Verified
Kimi K3—
MiniMax M380.5%
Source
Not directly comparable
SWE-bench Pro
Kimi K3—
MiniMax M359%
Source
Not directly comparable
Terminal-Bench 2.1
Kimi K3—
MiniMax M366.0%
Source
Not directly comparable
NL2Repo
Kimi K3—
MiniMax M342.1%
Source
Not directly comparable
VIBE V2
Kimi K3—
MiniMax M350.1%
Source
Not directly comparable
SVG-Bench
Kimi K3—
MiniMax M363.7%
Source
Not directly comparable
KernelBench Hard
Kimi K3—
MiniMax M328.8%
Source
Not directly comparable

Reasoning

ARC-AGI-1
Kimi K394.50%
Source
MiniMax M3—
Not directly comparable
ARC-AGI-2
Kimi K360.4%
Source
MiniMax M3—
Not directly comparable

Multimodal

OfficeQA Pro
Kimi K363.3%
Source
MiniMax M345.1%
Source
Kimi K3 leads this result
MMMU-Pro
Kimi K381.6%
Source
MiniMax M378.1%
Source
Kimi K3 leads this result
MMMU-Pro w/ Python
Kimi K383.4%
Source
MiniMax M3—
Not directly comparable
CharXiv w/o tools
Kimi K384.8%
Source
MiniMax M3—
Not directly comparable
CharXiv
Kimi K391.3%
Source
MiniMax M3—
Not directly comparable
MathVision
Kimi K394.3%
Source
MiniMax M3—
Not directly comparable
MathVision w/ Python
Kimi K397.8%
Source
MiniMax M3—
Not directly comparable
BabyVision w/ Python
Kimi K385.7%
Source
MiniMax M3—
Not directly comparable
ZeroBench
Kimi K323.0%
Source
MiniMax M3—
Not directly comparable
ZeroBench w/ Python
Kimi K341.0%
Source
MiniMax M3—
Not directly comparable
WorldVQA ForceAnswer
Kimi K351.0%
Source
MiniMax M3—
Not directly comparable
OmniDocBench
Kimi K391.1%
Source
MiniMax M3—
Not directly comparable
PerceptionBench
Kimi K358.5%
Source
MiniMax M3—
Not directly comparable
OmniDocBench 1.5
Kimi K3—
MiniMax M391.6%
Source
Not directly comparable
VideoMMMU
Kimi K3—
MiniMax M384.6%
Source
Not directly comparable
Video-MME (with subtitle)
Kimi K3—
MiniMax M385.4%
Source
Not directly comparable

Knowledge

GPQA
Kimi K393.5%
Source
MiniMax M3—
Not directly comparable
GPQA-D
Kimi K393.5%
Source
MiniMax M3—
Not directly comparable
HLE
Kimi K356%
Source
MiniMax M3—
Not directly comparable
HLE w/o tools
Kimi K343.5%
Source
MiniMax M3—
Not directly comparable
GPQA Diamond (Vals)
Kimi K392.9%
Source
MiniMax M392.7%
Source
Kimi K3 leads this result
MMLU-Pro (Vals)
Kimi K388.0%
Source
MiniMax M384.2%
Source
Kimi K3 leads this result

Instruction following

Gray Swan IPI (15 attempts)
Kimi K352.7%
Source
MiniMax M3—
Not directly comparable

Math

USAMO 2026
Kimi K3—
MiniMax M385.7%
Source
Not directly comparable

64 public results · 11 shared

Related comparisons

Compare API pricing Read the methodology Open the model selector

Watch Kimi K3 vs MiniMax M3

One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.

Read a sample issue

Join 2,000+ readers.

Last updated October 2, 2026

Kimi K3 vs MiniMax M3

Decision reading

Which one for your work

Coding work

Agentic work

Long documents

Chat turn cost

Cache-heavy agent loop cost

Repository review cost

Which one for a specific job

What is actually comparable

Shape of the matched evidence

Category results, on a stated basis

Agentic

Coding

Knowledge

Reasoning

Multimodal

Multilingual

Instruction following

Math

What each workload costs

Chat turn

Repository review

Cache-heavy agent loop

Specification differences

Context window

API model ID

Cached-input rate

Documented inputs

Documented outputs

Provider availability

Reasoning profile

Weight access

License

Release date

If you already use one of these models

Questions

Which is better, Kimi K3 or MiniMax M3?

Which is better for coding, Kimi K3 or MiniMax M3?

Which is better for agentic tasks, Kimi K3 or MiniMax M3?

Which costs less, Kimi K3 or MiniMax M3?

Which has the larger context window, Kimi K3 or MiniMax M3?

Benchmark evidence

Agentic

Coding

Reasoning

Multimodal

Knowledge

Instruction following

Math

Related comparisons

Watch Kimi K3 vs MiniMax M3