Kimi K3

MoonshotAI · On OpenRouter since July 16, 2026 · Pinned version id: moonshotai/kimi-k3-20260715

Price

$6.00

mixed / 1M tokens

Mixed price per 1M tokens: 3 input tokens to 1 output token.

in $3.00 · out $15.00 · Cached input $0.30

Models that think before they answer can use many more output tokens, so your real cost can be higher than the figure shown.

US dollars per million tokens, converted to your currency at the NBP rate.

Tasks, with the evidence

Each line names the measurement it rests on. A task that is not listed has no measurement or declared feature we can show.

Writing code Top 10%

Measured by a third party

Code: 1,346 · Rank 6 of 122, among scored models · Design Arena, read November 21, 2026

Cost of this job: $2.29

Hard questions and reasoning

Measured by a third party

Intelligence Index: 17.4 · Rank 48 of 66, among scored models · Artificial Analysis, read November 21, 2026

Cost of this job: $1.29

Models that score at least as high for less in this scenario: 38

Websites and apps Top 10%

Measured by a third party

Websites: 1,334 · Rank 5 of 127, among scored models · Top 10% · Design Arena, read November 21, 2026

UI components: 1,336 · Rank 11 of 118, among scored models · Top 10% · Design Arena, read November 21, 2026

Web apps (agent): 1,269 · Rank 9 of 47, among scored models · Top 25% · Design Arena, read November 21, 2026

Full-stack apps (agent): 1,286 · Rank 6 of 45, among scored models · Top 25% · Design Arena, read November 21, 2026

Mobile apps (agent): 1,253 · Rank 8 of 50, among scored models · Top 25% · Design Arena, read November 21, 2026

Android apps (agent): 1,253 · Rank 7 of 41, among scored models · Top 25% · Design Arena, read November 21, 2026

Cost of this job: $1.42

Models that score at least as high for less in this scenario: 2

Games and 3D Top 10%

Measured by a third party

Game development: 1,347 · Rank 10 of 121, among scored models · Top 10% · Design Arena, read November 21, 2026

3D scenes: 1,352 · Rank 13 of 114, among scored models · Top 25% · Design Arena, read November 21, 2026

Godot games (agent): 1,193 · Rank 12 of 34, among scored models · Design Arena, read November 21, 2026

Games (agent): 1,202 · Rank 4 of 25, among scored models · Design Arena, read November 21, 2026

Blender 3D (agent): 1,122 · Rank 9 of 25, among scored models · Design Arena, read November 21, 2026

Cost of this job: $1.42

Models that score at least as high for less in this scenario: 5

Slides and spreadsheets

Measured by a third party

PowerPoint files (agent): 1,223 · Rank 8 of 29, among scored models · Design Arena, read November 21, 2026

HTML slides (agent): 1,157 · Rank 5 of 25, among scored models · Design Arena, read November 21, 2026

Spreadsheets (agent): 1,052 · Rank 19 of 32, among scored models · Design Arena, read November 21, 2026

Cost of this job: $1.42

Models that score at least as high for less in this scenario: 3

Charts and graphics Top 10%

Measured by a third party

Data visualization: 1,347 · Rank 4 of 112, among scored models · Top 10% · Design Arena, read November 21, 2026

SVG graphics: 1,292 · Rank 8 of 84, among scored models · Top 10% · Design Arena, read November 21, 2026

ASCII art: 1,352 · Rank 3 of 68, among scored models · Top 10% · Design Arena, read November 21, 2026

Cost of this job: $1.42

Models that score at least as high for less in this scenario: 2

Long documents

Declared, not measured

Context window: 1M tokens as declared. It says how much fits, not how well the model uses it.

Cost of this job: $3.78

Images, files, audio, video in

Declared, not measured

Accepts: images and video. Declared by the provider through OpenRouter; we did not test it.

Cost of this job: $1.35Text-token prices only; images, audio and video may be billed separately.

The size of each job is our assumption; the prices are real.

  • Writing code: Requests: 40. Input tokens per request: 30,000. Output tokens per request: 1,500. Share of input served from cache: 70%.
  • Hard questions and reasoning: Requests: 10. Input tokens per request: 3,000. Output tokens per request: 8,000. Share of input served from cache: 0%.
  • Websites and apps; Games and 3D; Slides and spreadsheets; Charts and graphics: Requests: 8. Input tokens per request: 12,000. Output tokens per request: 10,000. Share of input served from cache: 30%.
  • Long documents: Requests: 3. Input tokens per request: 400,000. Output tokens per request: 4,000. Share of input served from cache: 0%.
  • Images, files, audio, video in: Requests: 15. Input tokens per request: 20,000. Output tokens per request: 3,000. Share of input served from cache: 30%.

Thinking is billed as output tokens; a model that thinks longer costs more than this estimate. Ranks are among all models we list, including older ones. The share is rounded up to whole models, so a rank slightly past the exact share still counts. Models with equal scores share the better rank.

Cheaper models with a similar benchmark score

Rule: an Intelligence Index of at least 14.4, a mixed price at least 25% lower, at least the same input types, and tool support. Lower mixed price (3 input tokens to 1 output token); check the flags and your own prompt sizes.

Models that pass this rule: 18. The cheapest of them are shown: 15.

  1. MiMo-V2.6-Flash Xiaomi 97% cheaper $0.175 · Index 26.7 (this model: 17.4) Differences to check: Lower max output Compare
  2. Muse Ember Meta 97% cheaper $0.20 · Index 36.0 (this model: 17.4) Compare
  3. Qwen3.8 Flash Qwen 96% cheaper $0.23 · Index 28.6 (this model: 17.4) Differences to check: Lower max output Compare
  4. GLM 5.3 Flash Z.ai 96% cheaper $0.2375 · Index 27.8 (this model: 17.4) Differences to check: Lower max output Compare
  5. MiniMax M3.1 Flash MiniMax 96% cheaper $0.2625 · Index 29.1 (this model: 17.4) Compare
Show more: 10
  1. MiniMax M3 MiniMax 91% cheaper $0.525 · Index 18.9 (this model: 17.4) Compare
  2. MiMo-V2.6-Pro Xiaomi 91% cheaper $0.5438 · Index 29.9 (this model: 17.4) Differences to check: Lower max output Compare
  3. Qwen3.8 27B Qwen 84% cheaper $0.9563 · Index 20.1 (this model: 17.4) Smaller context: 262k Differences to check: Lower max output Compare
  4. Step 5 Preview StepFun 76% cheaper $1.43 · Index 28.5 (this model: 17.4) Differences to check: Lower max output Compare
  5. Gemini 3.8 Flash Google 75% cheaper $1.50 · Index 30.8 (this model: 17.4) Differences to check: Lower max output Compare
  6. Muse Spark 1.3 Meta 67% cheaper $2.00 · Index 36.6 (this model: 17.4) Compare
  7. Muse Spark Open Meta 67% cheaper $2.00 · Index 24.9 (this model: 17.4) Compare
  8. Qwen4 Max Qwen 64% cheaper $2.14 · Index 42.9 (this model: 17.4) Differences to check: Lower max output Compare
  9. Gemini 4 Hydrogen Google 50% cheaper $3.00 · Index 53.9 (this model: 17.4) Compare
  10. Qwen3.8 Max (0902) Qwen 50% cheaper $3.00 · Index 33.7 (this model: 17.4) Differences to check: Lower max output Compare

Price by provider

The price used in the catalog is the vendor's own; when the vendor has no API of its own, it is the cheapest regular host.

ProviderInputOutputCached inputMixedNotes
Wafer $0.276$9.60 $0.13 $2.61 57% cheaper US · mxfp4 · Numeric precision used by this host. Lower precision (fp4, int4) is cheaper and can change quality.
Relace $0.1805$11.01 $0.18 $2.89 52% cheaper fp4 · Numeric precision used by this host. Lower precision (fp4, int4) is cheaper and can change quality.
Chutes $1.85$9.50 $0.185 $3.76 37% cheaper mxfp4 · Numeric precision used by this host. Lower precision (fp4, int4) is cheaper and can change quality.
Inference.net $0.245$15.00 $0.214 $3.93 34% cheaper fp4 · Numeric precision used by this host. Lower precision (fp4, int4) is cheaper and can change quality.
Inference.net $0.245$15.00 $0.214 $3.93 34% cheaper US · fp4 · Numeric precision used by this host. Lower precision (fp4, int4) is cheaper and can change quality.
KT $0.45$15.00 $0.30 $4.09 32% cheaper
Wafer $0.192$16.00 $0.182 $4.14 31% cheaper mxfp4 · Numeric precision used by this host. Lower precision (fp4, int4) is cheaper and can change quality.
Io Net $0.90$14.00 $0.15 $4.18 30% cheaper fp4 · Numeric precision used by this host. Lower precision (fp4, int4) is cheaper and can change quality.
Show all providers: 24
  • Sail Research: $0.345 / $16.58 · $4.40
  • DigitalOcean: $2.55 / $12.95 · $5.15
  • Makora: $1.00 / $18.00 · $5.25
  • DeepInfra: $2.85 / $14.25 · $5.70
  • MoonshotAI: $3.00 / $15.00 · $6.00
  • Alibaba: $3.00 / $15.00 · $6.00
  • Amazon Bedrock: $3.00 / $15.00 · $6.00
  • BaseTen: $3.00 / $15.00 · $6.00
  • Fireworks: $3.00 / $15.00 · $6.00
  • Modal: $3.00 / $15.00 · $6.00
  • Nebius: $3.00 / $15.00 · $6.00
  • Parasail: $3.00 / $15.00 · $6.00
  • River: $3.00 / $15.00 · $6.00
  • Together: $3.00 / $15.00 · $6.00
  • Amazon Bedrock: $3.30 / $16.50 · $6.60
  • Fireworks: $4.50 / $22.50 · $9.00

Other pricing modes

Specifications

Context window
1M
Max output
944k
Accepts
text, images, and video
Answers with
text
Tools
Yes
Structured output
Yes
Reasoning
Optional
Thinking effort levels: none, low, high, and max
Thinking is on by default
Weights on Hugging Face
Hugging Face A Hugging Face repository is linked. We did not check the licence; read it before you build on it.
Knowledge cutoff
No data
On OpenRouter since
July 16, 2026
Price check
one source

As listed by OpenRouter. "Not listed" means OpenRouter does not list it for this model.

Benchmark scores

Scores come from third parties and are relayed by OpenRouter. We do not run the tests. Higher is better.

SourceScoreRank among scored
Intelligence Index
Artificial Analysis, read November 21, 2026
17.4Rank 48 of 66, among scored models
Websites
Design Arena, read November 21, 2026
1,334Rank 5 of 127, among scored models Top 10%
UI components
Design Arena, read November 21, 2026
1,336Rank 11 of 118, among scored models Top 10%
Game development
Design Arena, read November 21, 2026
1,347Rank 10 of 121, among scored models Top 10%
3D scenes
Design Arena, read November 21, 2026
1,352Rank 13 of 114, among scored models Top 25%
Data visualization
Design Arena, read November 21, 2026
1,347Rank 4 of 112, among scored models Top 10%
SVG graphics
Design Arena, read November 21, 2026
1,292Rank 8 of 84, among scored models Top 10%
ASCII art
Design Arena, read November 21, 2026
1,352Rank 3 of 68, among scored models Top 10%
Code
Design Arena, read November 21, 2026
1,346Rank 6 of 122, among scored models Top 10%
Web apps (agent)
Design Arena, read November 21, 2026
1,269Rank 9 of 47, among scored models Top 25%
Full-stack apps (agent)
Design Arena, read November 21, 2026
1,286Rank 6 of 45, among scored models Top 25%
Mobile apps (agent)
Design Arena, read November 21, 2026
1,253Rank 8 of 50, among scored models Top 25%
Android apps (agent)
Design Arena, read November 21, 2026
1,253Rank 7 of 41, among scored models Top 25%
Godot games (agent)
Design Arena, read November 21, 2026
1,193Rank 12 of 34, among scored models
Games (agent)
Design Arena, read November 21, 2026
1,202Rank 4 of 25, among scored models
Blender 3D (agent)
Design Arena, read November 21, 2026
1,122Rank 9 of 25, among scored models
PowerPoint files (agent)
Design Arena, read November 21, 2026
1,223Rank 8 of 29, among scored models
HTML slides (agent)
Design Arena, read November 21, 2026
1,157Rank 5 of 25, among scored models
Spreadsheets (agent)
Design Arena, read November 21, 2026
1,052Rank 19 of 32, among scored models
agents/codetovideo
Design Arena, read November 21, 2026
1,274Rating 1,274; rank 4 at the source. Too few scored models for a tier.
agents/pdf
Design Arena, read November 21, 2026
1,276Rating 1,276; rank 6 at the source. Too few scored models for a tier.

Intelligence, Coding and Agentic indices: Artificial Analysis (artificialanalysis.ai). Design categories: Design Arena (designarena.ai). Both are relayed by the OpenRouter public model list and were read on November 21, 2026. We compute ranks, tiers and costs from these numbers; the sources have not reviewed them. Scores belong to the model, not to the route it is served on.

Artificial Analysis · Design Arena · Relayed by OpenRouter. opens in a new tab

Sources and dates

Prices: OpenRouter routes and a second price table, checked November 21, 2026. Specifications: OpenRouter, November 21, 2026. Exchange rate: NBP, November 20, 2026. Benchmarks: Artificial Analysis and Design Arena via OpenRouter.