NOUMENOpen app
NOUMEN

Moonshot · Open weights

Kimi K3 TEE

Added Jul 31, 2026Open weights

Compare models

Choose up to 3 models to compare.1 of 3 chosen

Loading models…

Use in the app

Kimi K3 is Moonshot AI's open-weight, always-thinking multimodal model for long-context reasoning, coding, tool use, and native image understanding. Running inside a TEE (Trusted Execution Environment), with provider attestation support.

  • Reasoning
  • Tools
  • Vision
  • Structured output
Input
Text, Images
Output
Text
Reasoning effort
low · high · max
Discovery tier
Open weights

Pricing

USD per 1M tokens
Input
$3.00
Output
$15.00
Cache read
$1.50

Reference prices as published by the model provider. API example

Specifications

Context window
1M
Max output
65.5K
Average output (7 days)
1.1K tokens
Longer than 75% of 473 models

Side by side

Kimi K3 TEE compared

Specifications, prices, speed and benchmark scores next to each other. Benchmarks below show every compared model too.

Compare models

Choose up to 3 models to compare.1 of 3 chosen

Loading models…

Kimi K3 TEEMoonshotThis modelKimi K2.6 ThinkingMoonshotRemove
ProviderMoonshotMoonshot
Discovery tierOpen weightsOpen weights
AddedJul 31, 2026Apr 16, 2026
Limits
Context window1.05MBest262.14K
Max output65.54K65.54KBest
Average output1.1K tokens4K tokens
Pricing, USD per 1M tokens
Input$3.00$0.45Best
Output$15.00$1.95Best
Cache read$1.50$0.045Best
Performance
Throughput87.4 tok/sBest40.8 tok/s
Time to first token3.2 s2.9 sBest
Request success (24h)98.0%99.8%Best
Benchmark indexes
Intelligence Index43.6Best27.0
Coding Index76.2Best61.8
Capabilities
ReasoningYesYes
ToolsYesYes
VisionYesYes
Structured outputYesNo
Parallel toolsNoYes
Reasoning effortlow, high, max—

Public evaluations

Benchmarks

Independent evaluation results, with how each score ranks among the models in the Noumen catalog.

  • Kimi K3 TEE
  • Kimi K2.6 Thinking

Intelligence Index

  • Kimi K3 TEE43.6
  • Kimi K2.6 Thinking27.0

Kimi K3 TEE tested at max reasoning effort

Kimi K3 TEE: Better than 89% of 326 scored models

3.8Catalog models by score55.1

Indexes

  • Coding Index

    Composite of the coding evaluations

    • Kimi K3 TEE: 76.2Top
    • Kimi K2.6 Thinking: 61.8

Agentic work

  • AutomationBench-AA

    Workflow automation, penalized for guardrail violations

    • Kimi K3 TEE: 58.3%Top
    • Kimi K2.6 Thinking: 13.0%

  • AutomationBench-AA tasks completed

    Workflows finished end to end without a guardrail violation

    • Kimi K3 TEE: 24.8%Top
    • Kimi K2.6 Thinking: 1.8%

  • Harvey LAB-AAv1.1

    Legal tasks passed in full, with no material hallucination

    • Kimi K3 TEE: 5.3%
    • Kimi K2.6 Thinking: —

  • AA-Briefcasev1.1

    Agentic knowledge work, rated head to head

    • Kimi K3 TEE: 1,499 EloTop
    • Kimi K2.6 Thinking: 819 Elo

  • GDPval-AAv2.1

    Economically valuable occupational tasks, rated head to head

    • Kimi K3 TEE: 1,534 EloTop
    • Kimi K2.6 Thinking: 1,038 Elo

Document reasoning

  • GDP.pdf

    Professional reasoning over PDF documents, all-pass rate

    • Kimi K3 TEE: 22.0%Top
    • Kimi K2.6 Thinking: 13.0%

  • AA-LCRv1.1

    Reasoning across long documents

    • Kimi K3 TEE: 88.7%Top
    • Kimi K2.6 Thinking: 81.0%

  • MLCR-AA

    Long-context reasoning over medical records

    • Kimi K3 TEE: 21.7%Top
    • Kimi K2.6 Thinking: 11.1%

Reasoning and knowledge

  • Humanity’s Last Exam

    Expert-level questions across academic fields

    • Kimi K3 TEE: 46.9%Top
    • Kimi K2.6 Thinking: 37.5%

  • GPQA Diamond

    Graduate-level science questions

    • Kimi K3 TEE: 93.5%Top
    • Kimi K2.6 Thinking: 91.1%

  • IFBench

    Following precise output instructions

    • Kimi K3 TEE: —
    • Kimi K2.6 Thinking: 76.0%

Coding

  • Terminal-Benchv4.0

    Practical software engineering in a terminal

    • Kimi K3 TEE: 12.6%Top
    • Kimi K2.6 Thinking: 0.5%

  • SciCode

    Scientific computing in Python

    • Kimi K3 TEE: 59.5%Top
    • Kimi K2.6 Thinking: 51.5%

Legacy evaluations

  • Terminal-Bench Hard

    Earlier terminal task suite

    • Kimi K3 TEE: —
    • Kimi K2.6 Thinking: 43.9%

  • τ²-Bench Telecom

    Tool use in a customer-support simulation

    • Kimi K3 TEE: —
    • Kimi K2.6 Thinking: 95.9%

  • AA-LCR (unversioned)

    Long-context reasoning, earlier grading

    • Kimi K3 TEE: 88.7%Top
    • Kimi K2.6 Thinking: 81.0%

Scores from Artificial Analysis, updated Oct 26, 2026. Kimi K3 TEE was tested at max reasoning effort; compare scores measured at the same effort.Artificial Analysis

Live routing data

Performance and reliability

Recent averages across requests routed for this model. Your own speed depends on prompt length, reasoning effort and load.

Throughput
87.4 tok/s
Faster than 57% of 256 models
Time to first token
3.2 s
Quicker than 48% of 256 models
Prompt cache hits
83.6%
About $1.75 per 1M input tokens with caching
Request success (24h)
98.0%
Reliable

Cache hits are the share of input tokens recently served from a prompt cache across all requests. They are not a forecast of your own hit rate.

Hourly request success · last 72 hoursLowest hour 61.6%
  • Healthy (98% or more)
  • Degraded (90–98%)
  • Down (under 90%)
  • No requests

Developers

Use it through the API

Noumen’s API is OpenAI-compatible: any Chat Completions client works with a base URL and a key.

Read the quickstart
curl https://www.noumen.xyz/api/v1/chat/completions \
  -H "Authorization: Bearer $NOUMEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "TEE/kimi-k3",
    "reasoning_effort": "low",
    "messages": [{ "role": "user", "content": "Explain this stack trace." }]
  }'
import os
from openai import OpenAI

client = OpenAI(base_url="https://www.noumen.xyz/api/v1", api_key=os.environ["NOUMEN_API_KEY"])
reply = client.chat.completions.create(
    model="TEE/kimi-k3",
    messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)

Moonshot

Other Moonshot models, starting with the same family.