Model A

Google

75.37/100

Estimated · Public rank #12

90% interval 63.9–86.9

Gemini 3.8 Flash vs Kimi K2.6

Updated September 2, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.

Model B

Kimi K2.6

Moonshot AI

59.18/100

Estimated · Public rank #84

90% interval 49.3–69.0

Decision reading

Gemini 3.8 Flash has the higher public score estimate, 75.37 versus 59.18, but the 90% score intervals overlap. Treat that as a lead, not a settled winner.

1 results are shared. Category rows based on different benchmark sets are marked directional and do not name a winner.

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.

Long documents
Prompts that approach the documented context limit
Gemini 3.8 Flash
Gemini 3.8 Flash has the larger documented context window.
Confidence: documented
Chat turn cost
1K fresh input + 500 output tokens
Gemini 3.8 Flash
Gemini 3.8 Flash 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 3.8 Flash
Gemini 3.8 Flash has the lower estimated token cost for this stated workload. Kimi K2.6 has no published cached-input rate, so cached tokens use its listed input rate.
Confidence: rate-fallback

Show secondary and unsupported calls

Repository review cost
50K fresh input + 3K output tokens
Gemini 3.8 Flash
Gemini 3.8 Flash has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
Coding work
Code generation, repair, and software-engineering tasks
Not enough matched evidence
No shared weighted benchmark basis supports a winner.
Confidence: limited
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
No shared weighted benchmark basis supports a winner.
Confidence: limited

What is actually comparable

Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.

11 Gemini 3.8 Flash only

1 Shared

31 Kimi K2.6 only

Shared results: 1
Gemini 3.8 Flash only: 11
Kimi K2.6 only: 31
Like-for-like categories: 0 / 8

Category results, on a stated basis

Each row states whether both averages use the same weighted benchmark set. Directional and not-comparable rows remain visible, but they never receive a winner in this template.

Agentic

Not comparable

Gemini 3.8 Flash: Not measured
Kimi K2.6: 73.5
Weighted basis: 0 vs 3 rows
Reading: Not comparable

Coding

Not comparable

Gemini 3.8 Flash: Not measured
Kimi K2.6: 64.4
Weighted basis: 0 vs 3 rows
Reading: Not comparable

Reasoning

Not comparable

Gemini 3.8 Flash: Not measured
Kimi K2.6: Not measured
Weighted basis: 0 vs 0 rows
Reading: Not comparable

Knowledge

Not comparable

Gemini 3.8 Flash: Not measured
Kimi K2.6: 42.2
Weighted basis: 0 vs 2 rows
Reading: Not comparable

Math

Not comparable

Gemini 3.8 Flash: Not measured
Kimi K2.6: 67.1
Weighted basis: 0 vs 4 rows
Reading: Not comparable

Multilingual

Not comparable

Gemini 3.8 Flash: Not measured
Kimi K2.6: Not measured
Weighted basis: 0 vs 0 rows
Reading: Not comparable

Multimodal

Not comparable

Gemini 3.8 Flash: Not measured
Kimi K2.6: 79.8
Weighted basis: 0 vs 2 rows
Reading: Not comparable

Instruction following

Not comparable

Gemini 3.8 Flash: Not measured
Kimi K2.6: Not measured
Weighted basis: 0 vs 0 rows
Reading: Not comparable

Category averages with the server-provided evidence basis for Gemini 3.8 Flash and Kimi K2.6
Category	Gemini 3.8 Flash	Kimi K2.6	Weighted basis	Reading
Agentic	Not measured	73.5	Not comparable0 vs 3 rows	Not comparable
Coding	Not measured	64.4	Not comparable0 vs 3 rows	Not comparable
Reasoning	Not measured	Not measured	Not comparable0 vs 0 rows	Not comparable
Knowledge	Not measured	42.2	Not comparable0 vs 2 rows	Not comparable
Math	Not measured	67.1	Not comparable0 vs 4 rows	Not comparable
Multilingual	Not measured	Not measured	Not comparable0 vs 0 rows	Not comparable
Multimodal	Not measured	79.8	Not comparable0 vs 2 rows	Not comparable
Instruction following	Not measured	Not measured	Not comparable0 vs 0 rows	Not comparable

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.

A shared-evidence shape is not available.

BenchLM does not draw a radar or infer missing axes when the matched evidence is too sparse.

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 3.8 Flash: $0.00262; Fits in one request
Kimi K2.6: $0.00295; Fits in one request

Gemini 3.8 Flash has the lower modeled cost

Costs use the listed standard API rates.

Repository review

50K fresh input + 3K output tokens

Gemini 3.8 Flash: $0.04875; Fits in one request
Kimi K2.6: $0.0595; Fits in one request

Gemini 3.8 Flash 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 3.8 Flash: $0.0675; Fits in one request
Kimi K2.6: $0.249; Fits in one request; Cached input priced at the published list-input rate

Gemini 3.8 Flash has the lower modeled cost

Kimi K2.6 has no published cached-input rate, so cached tokens use its listed input rate.

Specification differences

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

SpecificationGemini 3.8 FlashKimi K2.6

Context window

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

Gemini 3.8 Flash

Google Gemini 3.8 Flash API documentation

Kimi K2.6

256K

API model ID

Gemini 3.8 Flash

gemini-3.8-flash

Google Gemini 3.8 Flash API documentation

Kimi K2.6

Not sourced

Cached-input rate

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

Gemini 3.8 Flash

$0.075 per 1M cached input tokens

Google Gemini API pricing

Kimi K2.6

Not published

Documented inputs

Gemini 3.8 Flash

text, image, video, audio, pdf

Google Gemini 3.8 Flash API documentation

Kimi K2.6

Not sourced

Documented outputs

Gemini 3.8 Flash

text

Google Gemini 3.8 Flash API documentation

Kimi K2.6

Not sourced

Provider availability

Gemini 3.8 Flash

Generally Available · Gemini API, Google AI Studio, Android Studio, Stitch, Gemini app, Gemini Enterprise, Google AI Mode, Google Sheets, Google Antigravity

Google Gemini 3.8 Flash and Gemini 3.8 Flash Cyber launch

Kimi K2.6

Not sourced

Reasoning profile

Gemini 3.8 Flash

Reasoning

Kimi K2.6

Reasoning

Weight access

Gemini 3.8 Flash

Proprietary

Kimi K2.6

Open Weight

License

Gemini 3.8 Flash

Proprietary

Kimi K2.6

Open Weight

Release date

Gemini 3.8 Flash

2026-09-02

Kimi K2.6

2026-04-20

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: Gemini 3.8 Flash has the higher public score estimate, 75.37 versus 59.18, but the 90% score intervals overlap.
Workload cost: Repository review: $0.04875 vs $0.0595. Cache-heavy agent loop: $0.0675 vs $0.249.
Context tradeoff: Gemini 3.8 Flash has the larger documented window (1M).

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

Self-host vs API cost

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

Gemini 3.8 Flash

API / mo$3,375

Self-host / moNot listed

Break-even—

Proprietary model — self-hosting not applicable.

Kimi K2.6

API / mo$3,713

Self-host / mo$18,221

Break-even326M/day

Model the full break-even

Benchmark evidence

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

Browse raw public benchmark evidence43 rows

Agentic

Finance Agent v2
Gemini 3.8 Flash61.4%
Source
Kimi K2.6—
Not directly comparable
Terminal-Bench 2.1
Gemini 3.8 Flash89.4%
Source
Kimi K2.6—
Not directly comparable
Terminal-Bench 4.0
Gemini 3.8 Flash19.10%
Source
Kimi K2.6—
Not directly comparable
OSWorld 2.0
Gemini 3.8 Flash59.0%
Source
Kimi K2.64.6%
Source
Gemini 3.8 Flash leads this result
Terminal-Bench 2.0
Gemini 3.8 Flash—
Kimi K2.666.7%
Source
Not directly comparable
BrowseComp
Gemini 3.8 Flash—
Kimi K2.683.2%
Source
Not directly comparable
OSWorld-Verified
Gemini 3.8 Flash—
Kimi K2.673.1%
Source
Not directly comparable
Toolathlon
Gemini 3.8 Flash—
Kimi K2.650%
Source
Not directly comparable
MCP Atlas
Gemini 3.8 Flash—
Kimi K2.655.9%
Source
Not directly comparable
Claw-Eval
Gemini 3.8 Flash—
Kimi K2.662.3%
Source
Not directly comparable
DeepSearchQA
Gemini 3.8 Flash—
Kimi K2.692.5%
Source
Not directly comparable
WideResearch
Gemini 3.8 Flash—
Kimi K2.680.8%
Source
Not directly comparable
Gert Labs
Gemini 3.8 Flash—
Kimi K2.656.82%
Source
Not directly comparable
ResearchClawBench
Gemini 3.8 Flash—
Kimi K2.618.0%
Source
Not directly comparable

Coding

deepSwe
Gemini 3.8 Flash73.8%
Source
Kimi K2.6—
Not directly comparable
Terminal-Bench 2.1
Gemini 3.8 Flash89.4%
Source
Kimi K2.6—
Not directly comparable
SWE-bench Verified
Gemini 3.8 Flash—
Kimi K2.680.2%
Source
Not directly comparable
LiveCodeBench v6
Gemini 3.8 Flash—
Kimi K2.689.6%
Source
Not directly comparable
SWE-bench Pro
Gemini 3.8 Flash—
Kimi K2.658.6%
Source
Not directly comparable
SWE Multilingual
Gemini 3.8 Flash—
Kimi K2.676.7%
Source
Not directly comparable
SciCode
Gemini 3.8 Flash—
Kimi K2.652.2%
Source
Not directly comparable
Terminal-Bench 2.0
Gemini 3.8 Flash—
Kimi K2.666.7%
Source
Not directly comparable
Vibe Code Bench
Gemini 3.8 Flash—
Kimi K2.637.89%
Source
Not directly comparable
cursorBench31
Gemini 3.8 Flash—
Kimi K2.647.6%
Source
Not directly comparable

Knowledge

HLE-Verified
Gemini 3.8 Flash54.9%
Source
Kimi K2.6—
Not directly comparable
LABBench2
Gemini 3.8 Flash86.2%
Source
Kimi K2.6—
Not directly comparable
BioMysteryBench (human-solvable)
Gemini 3.8 Flash88.8%
Source
Kimi K2.6—
Not directly comparable
BioMysteryBench (human-difficult)
Gemini 3.8 Flash56.5%
Source
Kimi K2.6—
Not directly comparable
GPQA
Gemini 3.8 Flash—
Kimi K2.690.5%
Source
Not directly comparable
GPQA-D
Gemini 3.8 Flash—
Kimi K2.690.5%
Source
Not directly comparable
HLE
Gemini 3.8 Flash—
Kimi K2.634.7%
Source
Not directly comparable

Math

AIME26
Gemini 3.8 Flash—
Kimi K2.696.4%
Source
Not directly comparable
HMMT Feb 2026
Gemini 3.8 Flash—
Kimi K2.692.7%
Source
Not directly comparable
MMAnswerBench
Gemini 3.8 Flash—
Kimi K2.686.0%
Source
Not directly comparable
FrontierMath v2 (Tiers 1-3)
Gemini 3.8 Flash—
Kimi K2.638.966%
Source
Not directly comparable
FrontierMath v2 (Tier 4)
Gemini 3.8 Flash—
Kimi K2.614.580%
Source
Not directly comparable

Multimodal

CharXiv w/o tools
Gemini 3.8 Flash86.2%
Source
Kimi K2.6—
Not directly comparable
LVBench
Gemini 3.8 Flash87.1%
Source
Kimi K2.6—
Not directly comparable
MMMU-Pro
Gemini 3.8 Flash—
Kimi K2.679.4%
Source
Not directly comparable
MMMU-Pro w/ Python
Gemini 3.8 Flash—
Kimi K2.680.1%
Source
Not directly comparable
CharXiv
Gemini 3.8 Flash—
Kimi K2.680.4%
Source
Not directly comparable
MathVision
Gemini 3.8 Flash—
Kimi K2.687.4%
Source
Not directly comparable
V*
Gemini 3.8 Flash—
Kimi K2.696.9%
Source
Not directly comparable

Frequently asked questions

Which is better, Gemini 3.8 Flash or Kimi K2.6?

Gemini 3.8 Flash has the higher public score estimate, 75.37 versus 59.18, 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 3.8 Flash or Kimi K2.6?

The published evidence does not provide a shared weighted coding basis for both models, so BenchLM does not name a coding winner.

Which is better for agentic tasks, Gemini 3.8 Flash or Kimi K2.6?

The published evidence does not provide a shared weighted agentic tasks basis for both models, so BenchLM does not name a agentic tasks winner.

Which costs less, Gemini 3.8 Flash or Kimi K2.6?

For the stated presets, chat costs $0.00263 on Gemini 3.8 Flash and $0.00295 on Kimi K2.6; repository review costs $0.04875 and $0.0595; the cache-heavy agent loop costs $0.0675 and $0.249. Kimi K2.6 has no published cached-input rate, so cached tokens use its listed input rate.

Which has the larger context window, Gemini 3.8 Flash or Kimi K2.6?

Gemini 3.8 Flash has the larger documented context window: 1M, compared with 256K.

Related comparisons

Compare API pricing Read the methodology Open the model selector

Last updated September 2, 2026

Watch Gemini 3.8 Flash vs Kimi K2.6

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

Read a sample issue

Join 2,000+ readers.

Gemini 3.8 Flash vs Kimi K2.6

Which one for your work

Long documents

Chat turn cost

Cache-heavy agent loop cost

Repository review cost

Coding work

Agentic work

What is actually comparable

Category results, on a stated basis

Agentic

Coding

Reasoning

Knowledge

Math

Multilingual

Multimodal

Instruction following

Shape of the matched evidence

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

Self-host vs API cost

Benchmark evidence

Agentic

Coding

Knowledge

Math

Multimodal

Frequently asked questions

Which is better, Gemini 3.8 Flash or Kimi K2.6?

Which is better for coding, Gemini 3.8 Flash or Kimi K2.6?

Which is better for agentic tasks, Gemini 3.8 Flash or Kimi K2.6?

Which costs less, Gemini 3.8 Flash or Kimi K2.6?

Which has the larger context window, Gemini 3.8 Flash or Kimi K2.6?

Related comparisons

Watch Gemini 3.8 Flash vs Kimi K2.6