ScaledThought
← All models

Moonshot AI

Kimi K3

Live · US + CAtextimageKimi K3 License (custom)

Moonshot's 2.8T-parameter model, with a new hybrid linear-attention architecture and native vision.

Price per 1M tokens

Input

$3.00

Cached input

$0.30

Output

$15.00

Priced per model · Same in US and CA · Free tier included · Pricing

AA Index

44

#3 of 14
open models we serve

Specs

Model id
kimi-k3
Weights released
Jul 27, 2026
Parameters
2.8T total · 104B active
Architecture
MoE
Context
1M tokens (1,048,576)
Input
text, image
License
Kimi K3 License (custom)
Regions
us, ca

Benchmarks

  • Artificial Analysis Intelligence Index44
  • GPQA Diamond93.5%

Sources: huggingface.co · github.com · implicator.ai

Use it

Three lines of code, or zero.

OpenAI-compatible. Use the model id below, or auto to let the router decide when this model is worth it.

Quickstart→
import os
from openai import OpenAI

client = OpenAI(base_url=os.environ["ST_BASE_URL"], api_key=os.environ["ST_API_KEY"])
resp = client.chat.completions.create(
    model="kimi-k3",
    messages=[{"role": "user", "content": "Hello"}],
)

Opens your agent with the prompt typed. You press Enter.