Gemini 4 Argon vs Kimi K3

Updated September 30, 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 estimate, 71.34 versus 64.59, but the 90% score intervals overlap. Treat that as a lead, not a settled winner. 5 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.

Model A

Gemini 4 Argon

Google

64.59/100

Estimated · Public rank #32

90% interval 53.1–76.1

Model B

Kimi K3

Moonshot AI

71.34/100

Supported · Public rank #13

90% interval 68.0–74.7

Shared results: 5
Gemini 4 Argon only: 12
Kimi K3 only: 43
Like-for-like categories: 2 / 8

Estimated: Gemini 4 Argon · Supported: Kimi K3How 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
Gemini 4 Argon
Gemini 4 Argon leads on the public coding lane, 68.4 to 61.6, with Supported evidence for both models, although the 90% intervals overlap.
Confidence: limited
Chat turn cost
1K fresh input + 500 output tokens
Gemini 4 Argon
Gemini 4 Argon 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
Gemini 4 Argon
Gemini 4 Argon has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates

Show secondary and unsupported calls

Repository review cost
50K fresh input + 3K output tokens
Gemini 4 Argon
Gemini 4 Argon has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
Gemini 4 Argon is scored on Estimated evidence for agentic, so the reading is directional rather than like-for-like.
Confidence: limited
Long documents
Prompts that approach the documented context limit
Not enough matched evidence
A complete context comparison is not sourced.
Confidence: limited

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.

68.4Gemini 4 Argon61.6Kimi K3

Like-for-like · BenchAlign v5.7

Gemini 4 Argon leads the like-for-like coding row, although the 90% intervals overlap.

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.

AutomationBenchAgentic
Normalized gap 20.5
Gemini 4 Argon:51.3%
Kimi K3:30.8%
Gemini 4 Argon source Kimi K3 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 v5.7 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.

Coding

Like-for-like

Gemini 4 Argon: 68.4; Supported · #8/144
Kimi K3: 61.6; Supported · #17/144
Basis: BenchAlign v5.7 lane · 4 vs 14 public rows
Reading: Gemini 4 Argon leads · intervals overlap

Knowledge

Like-for-like

Gemini 4 Argon: 72.9; Supported · #11/170
Kimi K3: 68.3; Supported · #17/170
Basis: BenchAlign v5.7 lane · 1 vs 6 public rows
Reading: Gemini 4 Argon leads · intervals overlap

Agentic

Directional only

Gemini 4 Argon: 63.7; Estimated · #14/119
Kimi K3: 68.0; Supported · #8/119
Basis: BenchAlign v5.7 lane · 7 vs 12 public rows
Reading: Directional only

Reasoning

Not comparable

Gemini 4 Argon: 77.1; Unranked · 3 rankable rows
Kimi K3: 65.8; #18/27
Basis: Provisional lane · 0 vs 1 weighted rows
Reading: Not comparable

Multimodal

Not comparable

Gemini 4 Argon: Not ranked
Kimi K3: 89.4; #1/49
Basis: Provisional lane · 0 vs 3 weighted rows
Reading: Not comparable

Multilingual

Not comparable

Gemini 4 Argon: Not ranked
Kimi K3: Not ranked
Basis: Provisional lane · 0 vs 0 weighted rows
Reading: Not comparable

Instruction following

Not comparable

Gemini 4 Argon: Not ranked
Kimi K3: Not ranked
Basis: Provisional lane · 0 vs 0 weighted rows
Reading: Not comparable

Math

Not comparable

Gemini 4 Argon: Not ranked
Kimi K3: Not ranked
Basis: Provisional lane · 0 vs 0 weighted rows
Reading: Not comparable

Category scores with the server-provided evidence basis for Gemini 4 Argon and Kimi K3
Category	Gemini 4 Argon	Kimi K3	Basis	Reading
Coding	68.4Supported · #8/144	61.6Supported · #17/144	Like-for-likeBenchAlign v5.7 lane · 4 vs 14 public rows	Gemini 4 Argon leads · intervals overlap
Knowledge	72.9Supported · #11/170	68.3Supported · #17/170	Like-for-likeBenchAlign v5.7 lane · 1 vs 6 public rows	Gemini 4 Argon leads · intervals overlap
Agentic	63.7Estimated · #14/119	68.0Supported · #8/119	Directional onlyBenchAlign v5.7 lane · 7 vs 12 public rows	Directional only
Reasoning	77.1Unranked · 3 rankable rows	65.8#18/27	Not comparableProvisional lane · 0 vs 1 weighted rows	Not comparable
Multimodal	Not ranked	89.4#1/49	Not comparableProvisional lane · 0 vs 3 weighted rows	Not comparable
Multilingual	Not ranked	Not ranked	Not comparableProvisional lane · 0 vs 0 weighted rows	Not comparable
Instruction following	Not ranked	Not ranked	Not comparableProvisional lane · 0 vs 0 weighted rows	Not comparable
Math	Not ranked	Not ranked	Not comparableProvisional lane · 0 vs 0 weighted rows	Not comparable

Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign v5.7) 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

Gemini 4 Argon: $0.007; Fit state unavailable
Kimi K3: $0.0105; Fits in one request

Gemini 4 Argon has the lower modeled cost

Costs use the listed standard API rates.

Repository review

50K fresh input + 3K output tokens

Gemini 4 Argon: $0.13; Fit state unavailable
Kimi K3: $0.195; Fits in one request

Gemini 4 Argon has the lower modeled cost

Costs use the listed standard API rates.

Cache-heavy agent loop

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

Gemini 4 Argon: $0.16; Fit state unavailable
Kimi K3: $0.27; Fits in one request

Gemini 4 Argon 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.

SpecificationGemini 4 ArgonKimi K3

Context window

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

Gemini 4 Argon

Not sourced

Google Gemini 4 Argon announcement

Kimi K3

1.05M

API model ID

Gemini 4 Argon

Not sourced

Kimi K3

Not sourced

Cached-input rate

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

Gemini 4 Argon

$0.1 per 1M cached input tokens

Google Gemini 4 Argon announcement

Kimi K3

$0.3 per 1M cached input tokens

Documented inputs

Gemini 4 Argon

Not sourced

Kimi K3

Not sourced

Documented outputs

Gemini 4 Argon

Not sourced

Kimi K3

Not sourced

Provider availability

Gemini 4 Argon

Limited Access · Fairwind Program

Google Gemini 4 Argon announcement

Kimi K3

Not sourced

Reasoning profile

Gemini 4 Argon

Reasoning

Kimi K3

Reasoning

Weight access

Gemini 4 Argon

Proprietary

Kimi K3

Pending

License

Gemini 4 Argon

Proprietary

Kimi K3

Pending

Release date

Gemini 4 Argon

2026-09-30

Kimi K3

2026-07-16

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 estimate, 71.34 versus 64.59, but the 90% score intervals overlap.
Workload cost: Repository review: $0.13 vs $0.195. Cache-heavy agent loop: $0.16 vs $0.27.
Context tradeoff: A complete documented context comparison is not available.

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

Questions

Which is better, Gemini 4 Argon or Kimi K3?

Kimi K3 has the higher public score estimate, 71.34 versus 64.59, but the 90% score intervals overlap. The higher estimate is not a decisive winner because the uncertainty ranges overlap.

Which is better for coding, Gemini 4 Argon or Kimi K3?

Gemini 4 Argon leads the public coding lane, 68.4 to 61.6, with Supported evidence for both models, although the 90% intervals overlap.

Which is better for agentic tasks, Gemini 4 Argon or Kimi K3?

Kimi K3 scores higher for agentic tasks on the public lane, 68 to 63.7. Gemini 4 Argon is scored on Estimated evidence for agentic tasks, so the reading is directional rather than like-for-like. BenchLM does not name a winner for agentic tasks from a directional reading; read the shared benchmark rows directly and test the models on the same task set.

Which costs less, Gemini 4 Argon or Kimi K3?

For the stated presets, chat costs $0.007 on Gemini 4 Argon and $0.0105 on Kimi K3; repository review costs $0.13 and $0.195; the cache-heavy agent loop costs $0.16 and $0.27. Costs use the listed standard API rates.

Which has the larger context window, Gemini 4 Argon or Kimi K3?

A complete documented context-window comparison is not available.

Benchmark evidence

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

Browse raw public benchmark evidence60 rows

Agentic

AutomationBench
Gemini 4 Argon51.3%
Source
Kimi K330.8%
Source
Gemini 4 Argon leads this result
Finance Agent v2
Gemini 4 Argon65.4%
Source
Kimi K3—
Not directly comparable
Terminal-Bench 4.0
Gemini 4 Argon57.40%
Source
Kimi K3—
Not directly comparable
Agents' Last Exam
Gemini 4 Argon39.5%
Source
Kimi K3—
Not directly comparable
OSWorld 2.0
Gemini 4 Argon69.2%
Source
Kimi K3—
Not directly comparable
CWE-bench v1
Gemini 4 Argon68.0%
Source
Kimi K3—
Not directly comparable
Terminal-Bench-Science 0.1 (6x verifier timeout)
Gemini 4 Argon57.6%
Source
Kimi K3—
Not directly comparable
Terminal-Bench 2.1
Gemini 4 Argon—
Kimi K388.3%
Source
Not directly comparable
BrowseComp
Gemini 4 Argon—
Kimi K391.2%
Source
Not directly comparable
DeepSearchQA
Gemini 4 Argon—
Kimi K395.0%
Source
Not directly comparable
Toolathlon-Verified
Gemini 4 Argon—
Kimi K373.2%
Source
Not directly comparable
MCP Atlas
Gemini 4 Argon—
Kimi K384.2%
Source
Not directly comparable
JobBench
Gemini 4 Argon—
Kimi K352.9%
Source
Not directly comparable
APEX-Agents
Gemini 4 Argon—
Kimi K337.6%
Source
Not directly comparable
SpreadsheetBench 2
Gemini 4 Argon—
Kimi K334.8%
Source
Not directly comparable
DECK-Bench
Gemini 4 Argon—
Kimi K373.5%
Source
Not directly comparable
Terminal-Bench 2.1 (Vals)
Gemini 4 Argon—
Kimi K380.9%
Source
Not directly comparable
ApprenticeBench
Gemini 4 Argon—
Kimi K318%
Source
Not directly comparable

Coding

DeepSWE
Gemini 4 Argon77.9%
Source
Kimi K367.5%
Source
Gemini 4 Argon leads this result
FrontierSWE v2
Shared source
Gemini 4 Argon55.1%
Kimi K325.9%
Gemini 4 Argon leads this result
Vibe Code Bench
Gemini 4 Argon91.90%
Source
Kimi K3—
Not directly comparable
PostTrainBench v1.1
Gemini 4 Argon45.3%
Source
Kimi K332.0%
Source
Gemini 4 Argon leads this result
cursorBench32
Gemini 4 Argon—
Kimi K360.8%
Source
Not directly comparable
FrontierSWE
Gemini 4 Argon—
Kimi K381.2%
Source
Not directly comparable
ProgramBench
Gemini 4 Argon—
Kimi K377.8%
Source
Not directly comparable
Kimi Code Bench v2
Gemini 4 Argon—
Kimi K372.9%
Source
Not directly comparable
sweMarathon
Gemini 4 Argon—
Kimi K342%
Source
Not directly comparable
PostTrain Bench
Gemini 4 Argon—
Kimi K336.6%
Source
Not directly comparable
MLS-Bench Lite
Gemini 4 Argon—
Kimi K348.3%
Source
Not directly comparable
VulcanBench v3
Gemini 4 Argon—
Kimi K373.7%
Source
Not directly comparable
OpenHarmony Bench
Gemini 4 Argon—
Kimi K357.3%
Source
Not directly comparable
LiveCodeBench (Vals)
Gemini 4 Argon—
Kimi K387.2%
Source
Not directly comparable
SWE-bench (Vals)
Gemini 4 Argon—
Kimi K393.4%
Source
Not directly comparable

Reasoning

Graphwalks BFS 128K
Gemini 4 Argon99.7%
Source
Kimi K3—
Not directly comparable
GraphWalks BFS 256K–1M
Gemini 4 Argon84.2%
Source
Kimi K3—
Not directly comparable
ARC-AGI-1
Gemini 4 Argon—
Kimi K394.50%
Source
Not directly comparable
ARC-AGI-2
Gemini 4 Argon—
Kimi K360.4%
Source
Not directly comparable

Multimodal

Chartography (no tools)
Gemini 4 Argon71.6%
Source
Kimi K3—
Not directly comparable
LVBench
Gemini 4 Argon91.7%
Source
Kimi K3—
Not directly comparable
OfficeQA Pro
Gemini 4 Argon—
Kimi K363.3%
Source
Not directly comparable
MMMU-Pro
Gemini 4 Argon—
Kimi K381.6%
Source
Not directly comparable
MMMU-Pro w/ Python
Gemini 4 Argon—
Kimi K383.4%
Source
Not directly comparable
CharXiv w/o tools
Gemini 4 Argon—
Kimi K384.8%
Source
Not directly comparable
CharXiv
Gemini 4 Argon—
Kimi K391.3%
Source
Not directly comparable
MathVision
Gemini 4 Argon—
Kimi K394.3%
Source
Not directly comparable
MathVision w/ Python
Gemini 4 Argon—
Kimi K397.8%
Source
Not directly comparable
BabyVision w/ Python
Gemini 4 Argon—
Kimi K385.7%
Source
Not directly comparable
ZeroBench
Gemini 4 Argon—
Kimi K323.0%
Source
Not directly comparable
ZeroBench w/ Python
Gemini 4 Argon—
Kimi K341.0%
Source
Not directly comparable
WorldVQA ForceAnswer
Gemini 4 Argon—
Kimi K351.0%
Source
Not directly comparable
OmniDocBench
Gemini 4 Argon—
Kimi K391.1%
Source
Not directly comparable
PerceptionBench
Gemini 4 Argon—
Kimi K358.5%
Source
Not directly comparable

Knowledge

LABBench2
Gemini 4 Argon88.8%
Source
Kimi K3—
Not directly comparable
GPQA
Gemini 4 Argon—
Kimi K393.5%
Source
Not directly comparable
GPQA-D
Gemini 4 Argon—
Kimi K393.5%
Source
Not directly comparable
HLE
Gemini 4 Argon—
Kimi K356%
Source
Not directly comparable
HLE w/o tools
Gemini 4 Argon—
Kimi K343.5%
Source
Not directly comparable
GPQA Diamond (Vals)
Gemini 4 Argon—
Kimi K392.9%
Source
Not directly comparable
MMLU-Pro (Vals)
Gemini 4 Argon—
Kimi K388.0%
Source
Not directly comparable

Instruction following

Gray Swan IPI (15 attempts)
Shared source
Gemini 4 Argon0.7%
Kimi K352.7%
Gemini 4 Argon leads this result

60 public results · 5 shared

Related comparisons

Compare API pricing Read the methodology Open the model selector

Watch Gemini 4 Argon vs Kimi K3

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 September 30, 2026

Gemini 4 Argon vs Kimi K3

Decision reading

Which one for your work

Coding work

Chat turn cost

Cache-heavy agent loop cost

Repository review cost

Agentic work

Long documents

Which one for a specific job

What is actually comparable

Shape of the matched evidence

Category results, on a stated basis

Coding

Knowledge

Agentic

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, Gemini 4 Argon or Kimi K3?

Which is better for coding, Gemini 4 Argon or Kimi K3?

Which is better for agentic tasks, Gemini 4 Argon or Kimi K3?

Which costs less, Gemini 4 Argon or Kimi K3?

Which has the larger context window, Gemini 4 Argon or Kimi K3?

Benchmark evidence

Agentic

Coding

Reasoning

Multimodal

Knowledge

Instruction following

Related comparisons

Watch Gemini 4 Argon vs Kimi K3