Model profile
Kimi
New in 2026

MoonshotAI: Kimi K3

MoonshotAI: Kimi K3 is a mid-range-priced multimodal generalist from Kimi with heavy runtime profile, large context posture, and the clearest fit around long-context research / agent workflows.

Best for: Long-context research / Agent workflowsHeavy latencyLarge contextMid-range pricing
Intelligence
48.3

Benchmark blend

Coding
72.0

Dev workflow signal

Context
1049K Tokens

Large

Input Price
$3.00

Mid-range tier

Decision snapshot
59

MoonshotAI: Kimi K3 currently reads as a mid-range multimodal option with large context and a heavy runtime profile.

Overall profile
Selective fit
Best for
Long-context research / Agent workflows
Latency tier
Heavy
Price tier
Mid-range
Source coverage
OpenRouterArtificial AnalysisVision signal

Decision Strip

Decision rail before the raw tables

Core buy-side signals stay in one pass. The rest of the page expands only after intelligence, speed, context, and price are clear.

Intelligence
48.3
48

General reasoning and benchmark headroom.

Situational
Speed
36 tok/s
27

TTFT 2.99s

Limited
Context
1049K Tokens
100

How much prompt and task state can stay in view.

Above average
Price
$3.00
62

$15.00 output / 1M

Competitive

Editorial Profile

MoonshotAI: Kimi K3 in one narrative

Positioning, tradeoffs, and fit are consolidated into one read instead of repeating the same story across separate cards.

Selective fitCoding score 72Math score N/AVision enabled

Kimi K3 is a 2.8T parameter open-weight multimodal reasoning model from Moonshot AI. It is suited for complex coding, knowledge work, and long-horizon agentic workflows, and is particularly strong at...

Identity

Kimi multimodal profile

Positioning

Long-context research / Agent workflows with large context and heavy runtime.

Cost posture

Balanced spend profile. Easier to justify in mixed production and exploration workloads.

Strengths
  • Coding indicators point to a strong developer workflow fit.

  • Large context headroom supports repo-wide prompts and long research sessions.

  • Vision-capable routing opens up multimodal review and extraction workflows.

Tradeoffs
  • Costs look manageable, but still deserve attention in always-on agents or batch jobs.

  • Latency profile is better for deliberate runs than rapid back-and-forth chat.

Best fit
  • Code generation, refactors, test writing, and tool-assisted debugging.

  • Image-grounded review, multimodal extraction, and UI audit workflows.

  • Long-context summarization, repo analysis, and policy or document review.

Explore Next

Similar profiles worth opening next

Kimi

Kimi K3 (max)

Intelligence
59.7
Context
N/A
Input Price
$3.00
Kimi

MoonshotAI: Kimi K2.6

Intelligence
45.1
Context
262K Tokens
Input Price
$0.95
Kimi

MoonshotAI: Kimi K2.7 Code

Intelligence
43.0
Context
262K Tokens
Input Price
$0.95

Benchmarks

Grouped by job-to-be-done

Only benchmark categories with actual signal are shown. Secondary values stay as simple definitions instead of nested micro-cards.

General intelligence

Broad reasoning, knowledge depth, and flagship benchmark posture.

Intelligence Index
48.3
GPQA
84.2%
HLE
25.0%
Coding

Software implementation, debugging quality, and coding benchmark signal.

Coding Index
72.0
SciCode
51.2%
Agent / tool use

Long-horizon execution quality and interactive benchmark evidence.

LCR
77.0%

Extra benchmark cuts are not available for this category yet.

Specs & Pricing

Technical snapshot and cost posture

Specs stay neutral, pricing gets emphasis through values rather than extra containers. Raw provider internals remain in metadata at the end.

Technical snapshot
Context Window
1049K Tokens
Vision
Enabled
Modalities
image, text, video
Tokenizer
Other
Max Completion
943718
Moderation
No
Supported Parameters
frequency_penaltyinclude_reasoninglogit_biaslogprobsmax_tokensmin_ppresence_penaltyreasoningreasoning_effortrepetition_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_ktop_logprobstop_p
Input Modalities
imagetextvideo
Output Modalities
text
Price architecture
Input
per 1M input tokens
$3.00
Output
per 1M output tokens
$15.00
Blended
AA 3:1 mix
$6.00

This model sits in a balanced spend range. It is easier to justify across both production and exploratory workflows.

OR Cache Read
$0.00

Metadata

Raw source tables at the end

Verification details remain available, but the page no longer forces them ahead of the editorial read.