Enterprise DNAEnterprise DNA
Moonshot AI logo

Models · Best coding

Kimi K3

by Moonshotai

Kimi K3 is an ultra-large-scale, open-weight multimodal reasoning model from Moonshot AI.

Visit Moonshotai ↗

Our verdict

Not reviewed yet

We have not reviewed Kimi K3 yet

No one at Enterprise DNA has checked this listing against its sources. The description comes from public listings gathered automatically. Any fact below shows where it came from and when it was last checked.

How we review records

From the listing

About Kimi K3

Kimi K3 is an ultra-large-scale, open-weight multimodal reasoning model from Moonshot AI.

Best for

Multimodal knowledge and code work

Notes

Kimi K3 is listed in the EDNA Models directory from the OpenRouter catalogue.

Pros

  • Planning a code change from a visual specification
  • text, image, video input; text output on the listed OpenRouter route

Cons

  • Assuming hosted access inherits every open-weight capability
  • Treating a product demo as an independent benchmark
  • Base text-token rates exclude other modalities, tools, caching, retries and any hosting-specific charges. Check actual usage before comparing cost per successful task.

Key facts

What we have checked

Input price
$0.192 per million tokens
Checked OpenRouter models API
Output price
$15.00 per million tokens
Checked OpenRouter models API
Price on akashml
$1.20 per million cached input tokens
Checked OpenRouter models API
Price on akashml
$1.20 per million input tokens
Checked OpenRouter models API
Price on akashml
$14.00 per million output tokens
Checked OpenRouter models API
Price on amazon-bedrock
$0.300 per million cached input tokens
Checked OpenRouter models API
Price on amazon-bedrock
$3.00 per million input tokens
Checked OpenRouter models API
Price on amazon-bedrock
$15.00 per million output tokens
Checked OpenRouter models API
Price on baseten
$0.300 per million cached input tokens
Checked OpenRouter models API
Gpqa diamond
93.1% (for 16 July 2026)
Checked Epoch AI (CC BY 4.0)
Frontiermath tiers 1 3 v2 private
72.2% (for 16 July 2026)
Checked Epoch AI (CC BY 4.0)
Simpleqa verified
50.6% (for 16 July 2026)
Checked Epoch AI (CC BY 4.0)

Each value shows when we last confirmed it and where it came from. Values we have not confirmed are left out, never guessed.

API access

How to call it

Call via OpenRouter with model id `moonshotai/kimi-k3`.

  • openrouter · moonshotai/kimi-k3 · Docs

OpenRouter id: moonshotai/kimi-k3

Price

What Kimi K3 costs, over time

Input price

$3.00 to $0.192

US dollars per million input tokens. 3 Sept 2026 to 31 Oct 2026, 38 days.

Source: OpenRouter models API. Checked .

Output price

$15.00 to $15.00

US dollars per million output tokens. 3 Sept 2026 to 31 Oct 2026, 38 days.

Source: OpenRouter models API. Checked .

Prices by host

HostPriceOver timeChecked
akashmlCached input$1.20 per million cached input tokensSince 24 Oct 2026No change recordedOpenRouter models API
akashmlInput$1.20 per million input tokensSince 24 Oct 2026No change recordedOpenRouter models API
akashmlOutput$14.00 per million output tokensSince 24 Oct 2026No change recordedOpenRouter models API
amazon-bedrockCached input$0.300 per million cached input tokensSince 24 Oct 2026No change recordedOpenRouter models API
amazon-bedrockInput$3.00 per million input tokensSince 24 Oct 2026No change recordedOpenRouter models API
amazon-bedrockOutput$15.00 per million output tokensSince 24 Oct 2026No change recordedOpenRouter models API
basetenCached input$0.300 per million cached input tokensSince 24 Oct 2026No change recordedOpenRouter models API
basetenInput$3.00 per million input tokensSince 24 Oct 2026No change recordedOpenRouter models API
basetenOutput$15.00 per million output tokensSince 24 Oct 2026No change recordedOpenRouter models API
chutesCached input$0.185 per million cached input tokensSince 24 Oct 2026No change recordedOpenRouter models API
chutesInput$1.85 per million input tokensSince 24 Oct 2026No change recordedOpenRouter models API
chutesOutput$9.50 per million output tokensSince 24 Oct 2026No change recordedOpenRouter models API
decartCached input$0.195 per million cached input tokensSince 30 Oct 20261 changeOpenRouter models API
decartInput$1.95 per million input tokensSince 30 Oct 20261 changeOpenRouter models API
decartOutput$9.75 per million output tokensSince 30 Oct 20261 changeOpenRouter models API
deepinfraCached input$0.285 per million cached input tokensSince 24 Oct 2026No change recordedOpenRouter models API
deepinfraInput$2.85 per million input tokensSince 24 Oct 2026No change recordedOpenRouter models API
deepinfraOutput$14.25 per million output tokensSince 24 Oct 2026No change recordedOpenRouter models API
digitaloceanCached input$0.255 per million cached input tokensSince 24 Oct 2026No change recordedOpenRouter models API
digitaloceanInput$2.55 per million input tokensSince 24 Oct 2026No change recordedOpenRouter models API
digitaloceanOutput$12.95 per million output tokensSince 24 Oct 2026No change recordedOpenRouter models API
fireworksCached input$0.210 per million cached input tokensSince 24 Oct 2026No change recordedOpenRouter models API
fireworksInput$2.10 per million input tokensSince 24 Oct 2026No change recordedOpenRouter models API
fireworksOutput$10.50 per million output tokensSince 24 Oct 2026No change recordedOpenRouter models API
inference-netCached input$0.195 per million cached input tokensSince 31 Oct 20262 changesOpenRouter models API
inference-netInput$0.234 per million input tokensSince 31 Oct 20262 changesOpenRouter models API
inference-netOutput$15.00 per million output tokensSince 1 Nov 20263 changesOpenRouter models API
makoraCached input$0.200 per million cached input tokensSince 29 Oct 20261 changeOpenRouter models API
makoraInput$0.500 per million input tokensSince 29 Oct 20261 changeOpenRouter models API
makoraOutput$13.00 per million output tokensSince 29 Oct 20261 changeOpenRouter models API
modalCached input$0.300 per million cached input tokensSince 24 Oct 2026No change recordedOpenRouter models API
modalInput$3.00 per million input tokensSince 24 Oct 2026No change recordedOpenRouter models API
modalOutput$15.00 per million output tokensSince 24 Oct 2026No change recordedOpenRouter models API
moonshotaiCached input$0.300 per million cached input tokensSince 24 Oct 2026No change recordedOpenRouter models API
moonshotaiInput$3.00 per million input tokensSince 24 Oct 2026No change recordedOpenRouter models API
moonshotaiOutput$15.00 per million output tokensSince 24 Oct 2026No change recordedOpenRouter models API
morphCached input$0.360 per million cached input tokensSince 29 Oct 20264 changesOpenRouter models API
morphInput$1.28 per million input tokensSince 29 Oct 20264 changesOpenRouter models API
morphOutput$12.00 per million output tokensSince 1 Nov 20264 changesOpenRouter models API
parasailCached input$0.300 per million cached input tokensSince 24 Oct 2026No change recordedOpenRouter models API
parasailInput$3.00 per million input tokensSince 24 Oct 2026No change recordedOpenRouter models API
parasailOutput$15.00 per million output tokensSince 24 Oct 2026No change recordedOpenRouter models API
pareto-alphaCached input$0.300 per million cached input tokensSince 24 Oct 2026No change recordedOpenRouter models API
pareto-alphaInput$3.00 per million input tokensSince 24 Oct 2026No change recordedOpenRouter models API
pareto-alphaOutput$15.00 per million output tokensSince 24 Oct 2026No change recordedOpenRouter models API
phalaCached input$0.255 per million cached input tokensSince 24 Oct 2026No change recordedOpenRouter models API
phalaInput$2.55 per million input tokensSince 24 Oct 2026No change recordedOpenRouter models API
phalaOutput$12.75 per million output tokensSince 24 Oct 2026No change recordedOpenRouter models API
qwenCached input$0.345 per million cached input tokensSince 24 Oct 2026No change recordedOpenRouter models API
qwenInput$3.45 per million input tokensSince 24 Oct 2026No change recordedOpenRouter models API

Usage

How much Kimi K3 is used

Tokens served a day

192.2B to 189.8B

tokens a day on OpenRouter. 19 Sept 2026 to 31 Oct 2026, 43 readings.

Source: OpenRouter model activity (735 of 735 series). Checked .

Scores

How Kimi K3 scores

Each score comes from the named source and shows when we last confirmed it. A flag means the score needs care.

ScoreValueSourceChecked
Artificial Analysis Intelligence Index17.4indexAs of 31 Oct 2026Artificial Analysis via OpenRouter
Artificial Analysis Coding Index76.2indexAs of 31 Oct 2026Artificial Analysis via OpenRouter
Artificial Analysis Agentic Index50indexAs of 31 Oct 2026Artificial Analysis via OpenRouter
LMArena text rating (Elo)1,476EloAs of 31 Oct 2026LMArena (CC-BY-4.0), published 2026-10-28
LMArena web development rating (Elo)1,656EloAs of 31 Oct 2026LMArena (CC-BY-4.0), published 2026-10-28
Design Arena rating (Elo)1,381EloAs of 31 Oct 2026Design Arena via OpenRouter
Apex agents (best-across-versions)47.8percentAs of 16 July 2026Epoch AI (CC BY 4.0)
Arc agi (best-across-versions)94.5percentAs of 16 July 2026Epoch AI (CC BY 4.0)
Arc agi 2 (best-across-versions)60.4percentAs of 16 July 2026Epoch AI (CC BY 4.0)
Chess puzzles (best-across-versions)37percentAs of 16 July 2026Epoch AI (CC BY 4.0)
Dtbench (best-across-versions)91.2percentAs of 16 July 2026Epoch AI (CC BY 4.0)
Frontiercode (best-across-versions)44.2percentAs of 16 July 2026Epoch AI (CC BY 4.0)
Show 15 more scores
ScoreValueSourceChecked
Frontiermath tier 4 v2 private (best-across-versions)39percentAs of 16 July 2026Epoch AI (CC BY 4.0)
Frontiermath tiers 1 3 v2 private (best-across-versions)72.2percentAs of 16 July 2026Epoch AI (CC BY 4.0)
Frontierswe (best-across-versions)25.9percentAs of 16 July 2026Epoch AI (CC BY 4.0)
Furniture assembly (best-across-versions)34.2percentAs of 16 July 2026Epoch AI (CC BY 4.0)
Gpqa diamond (best-across-versions)93.1percentAs of 16 July 2026Epoch AI (CC BY 4.0)
Lmca (best-across-versions)52.7percentAs of 16 July 2026Epoch AI (CC BY 4.0)
Mystery game puzzles (best-across-versions)26percentAs of 16 July 2026Epoch AI (CC BY 4.0)
Otis mock aime 2024 2025 (best-across-versions)Saturated97.2percentAs of 16 July 2026Epoch AI (CC BY 4.0)
Posttrainbench (best-across-versions)32percentAs of 16 July 2026Epoch AI (CC BY 4.0)
Proofbench (best-across-versions)87percentAs of 16 July 2026Epoch AI (CC BY 4.0)
Simplebench (best-across-versions)60.7percentAs of 16 July 2026Epoch AI (CC BY 4.0)
Simpleqa verified (best-across-versions)50.6percentAs of 16 July 2026Epoch AI (CC BY 4.0)
Surface evolver bench (best-across-versions)95percentAs of 16 July 2026Epoch AI (CC BY 4.0)
Weirdml (best-across-versions)82.6percentAs of 16 July 2026Epoch AI (CC BY 4.0)
Weirdml v3 (best-across-versions)15.4percentAs of 16 July 2026Epoch AI (CC BY 4.0)
Saturated
Top models score close to the maximum, so this test no longer separates them.

Compare

Compare Kimi K3 side by side

Sources

Where this comes from

A source shows what was published when we checked it. It does not certify security, performance or a working installation. Read our method.

Changelog

What changed

  1. Price changed

    cached input per mtok: 0.15 to 0.18

    Source: OpenRouter models API

  2. Price changed

    output per mtok: 10.5478 to 9.3388

    Source: OpenRouter models API

  3. Price changed

    input per mtok: 0.1558 to 0.1824

    Source: OpenRouter models API

  4. Price changed

    output per mtok: 14.626 to 12

    Source: OpenRouter models API

  5. Price changed

    output per mtok: 6.975 to 15

    Source: OpenRouter models API

  6. Price changed

    output per mtok: 9.94 to 9.6

    Source: OpenRouter models API

  7. Price changed

    cached input per mtok: 0.115 to 0.07475

    Source: OpenRouter models API

  8. Price changed

    output per mtok: 11 to 7.15

    Source: OpenRouter models API

  9. Price changed

    input per mtok: 0.23 to 0.1495

    Source: OpenRouter models API

  10. Price changed

    output per mtok: 9.75 to 10.5478

    Source: OpenRouter models API

  11. Price changed

    input per mtok: 0.1529 to 0.1558

    Source: OpenRouter models API

  12. Price changed

    cached input per mtok: 0.15 to 0.195

    Source: OpenRouter models API

  13. Price changed

    output per mtok: 7.5 to 6.975

    Source: OpenRouter models API

  14. Price changed

    input per mtok: 0.18 to 0.234

    Source: OpenRouter models API

  15. Price changed

    cached input per mtok: 0.329 to 0.13

    Source: OpenRouter models API

  16. Price changed

    output per mtok: 11.9 to 9.94

    Source: OpenRouter models API

  17. Price changed

    input per mtok: 0.33 to 0.276

    Source: OpenRouter models API

  18. Price changed

    cached input per mtok: 0.18 to 0.15

    Source: OpenRouter models API

  19. Price changed

    output per mtok: 9.3388 to 9.75

    Source: OpenRouter models API

  20. Price changed

    input per mtok: 0.1824 to 0.1529

    Source: OpenRouter models API

In the AI Pulse

Recent daily briefs from the Enterprise DNA AI Pulse that mention this. The short daily read on what is actually happening in AI.

See today's AI Pulse →

USAGE / OPENROUTER

How much Kimi K3 is used

Tokens processed through OpenRouter each day. It shows what developers actually run, not every company's private traffic.

Tokens, last 7 days
1.7T ▼16.6%
Requests, last 7 days
32.1M
Tool calls, last 7 days
18.5M
Intelligence score
17.4
People's vote
1476
Average cost per request
$2.48
Price per million tokens
$0.192 in · $15 out
0250B500B19 Sept: 192B tokens, 3.2M requests20 Sept: 173B tokens, 2.8M requests21 Sept: 218B tokens, 3.3M requests22 Sept: 193B tokens, 3.1M requests23 Sept: 189B tokens, 3.6M requests24 Sept: 196B tokens, 3.6M requests25 Sept: 211B tokens, 3.5M requests26 Sept: 185B tokens, 2.9M requests27 Sept: 203B tokens, 3.3M requests28 Sept: 229B tokens, 3.7M requests29 Sept: 276B tokens, 3.8M requests30 Sept: 264B tokens, 4.7M requests1 Oct: 204B tokens, 3.5M requests2 Oct: 212B tokens, 3.4M requests3 Oct: 211B tokens, 3M requests4 Oct: 225B tokens, 3.6M requests5 Oct: 299B tokens, 4.8M requests6 Oct: 282B tokens, 4M requests7 Oct: 286B tokens, 4.2M requests8 Oct: 329B tokens, 5.7M requests9 Oct: 339B tokens, 5.1M requests10 Oct: 307B tokens, 5M requests11 Oct: 309B tokens, 4.5M requests12 Oct: 452B tokens, 5.6M requests13 Oct: 497B tokens, 7.4M requests14 Oct: 394B tokens, 6.9M requests15 Oct: 236B tokens, 4.2M requests16 Oct: 203B tokens, 3.4M requests17 Oct: 180B tokens, 3.3M requests18 Oct: 250B tokens, 3.5M requests19 Oct: 283B tokens, 4.4M requests20 Oct: 279B tokens, 5.3M requests21 Oct: 285B tokens, 5.2M requests22 Oct: 413B tokens, 6.5M requests23 Oct: 342B tokens, 5M requests24 Oct: 189B tokens, 3.3M requests25 Oct: 198B tokens, 3.4M requests26 Oct: 197B tokens, 3.5M requests27 Oct: 300B tokens, 4.8M requests28 Oct: 316B tokens, 5.4M requests29 Oct: 273B tokens, 5.4M requests30 Oct: 229B tokens, 4.7M requests31 Oct: 190B tokens, 4.8M requests19 Sept3 Oct17 Oct31 OctSee where it ranks on the leaderboard →

MODEL FIELD GUIDE / SOURCE EVIDENCE

What you can do with Kimi K3.

Specifications reviewed; no EDNA task evaluation has been run. Sources checked 2026-10-30. Hosting routes and app features can differ.

DeveloperMoonshot AI
Exact model identifiermoonshotai/kimi-k3
Inputtext, image, video
Outputtext
Tool calling
Listed support
Structured output
Listed support
Reasoning controls
Listed support
Availability
available
Price basis
token
Licence verification
kimi-k3, open weights
Context length
1M tokens
Knowledge cutoff
Not established
Developer release date
Not established
First seen in the catalogue
16 Jul 2026
Listed retirement date
Not established

Base text-token rates exclude other modalities, tools, caching, retries and any hosting-specific charges. Check actual usage before comparing cost per successful task.

Consider it for

  • Planning a code change from a visual specification
  • Working through a document with explicit citations

Watch for

  • Assuming hosted access inherits every open-weight capability
  • Treating a product demo as an independent benchmark

PRACTICAL EXERCISE / NOT RUN BY ENTERPRISE DNA

Try the model on a small, checkable task.

Given a UI screenshot and acceptance rule that totals include tax, list visible controls, infer no hidden features, and propose tests for a tax-inclusive total.

Check the result: Observations and proposed tests are separate; no implementation or browser action is claimed.

Compare another approach

anthropic claude sonnet 5 5 ↗openai gpt 6 1 sol ↗
Inspect the sources and their scope
provider source ↗

Developer documentation for model family or task. OpenRouter route-specific properties remain catalogue claims; EDNA has not executed a task evaluation.

Checked 2026-10-19
catalogue source ↗

moonshotai/kimi-k3: listed route capabilities, availability and prices; provider endpoints can differ.

Checked 2026-10-30
openrouter.ai ↗

Source for reasoning controls, first seen, tool calling, context length, input, output, structured output.

Checked 2026-10-30
Licence source ↗

Licence declared in the moonshotai/Kimi-K3 model repository on Hugging Face, which holds the weights files. Read the licence terms before use.

Checked 2026-10-31
How to evaluate a model for your work →

FROM SPECIFICATIONS TO YOUR WORK

Is this the right model for your job?

Start with a realistic sample. These workflows match the model’s listed capabilities, not a verified result on your task.

Build an appWorkflow & starter prompt ↗Extract document dataWorkflow & starter prompt ↗Run a reliable agentWorkflow & starter prompt ↗

What would it cost?

Pick a common job and how many times a month you run it. Token sizes are our assumption for a typical run, shown under the result.

Working out the cost...

Token prices only. Per-request, image, tool, storage and subscription fees are not included. One token is about three quarters of an English word. Real counts vary by model and language. How tokens are counted ↗

Check the source

Price date: 31 Oct 2026, from OpenRouter models API. Prices are USD API token rates. Your app subscription, uploads, browsing and tools are separate.

Model source & availability ↗

Understand the evidence

Listed capabilities and third-party benchmark scores do not establish Enterprise DNA task-test results. Validate accuracy and total cost on your own sample.

Intelligence Index: 17.4 (2026-10-30) ↗Coding Index: 76.2 (2026-10-30) ↗Agentic Index: 50 (2026-10-30) ↗
Compare this model with alternatives →
Free 27-page guide

Get the free Developer’s Field Guide

A 27-page field guide to the AI coding workflow with Claude. Claude Code, MCP servers, the prompt patterns that work, and what to delegate. Free.

Enter your work email. We send it straight over, plus a few short notes worth knowing. Unsubscribe any time.

No spam. Unsubscribe any time.

← Back to ModelsSubmit your own entry →