AELDER
Moonshot AIModel API

Kimi K3 API

Pay 75% less.

On input and output tokens, compared with official prices.

OpenAI-compatible · Pay as you go

MODEL IDkimi-k3
One key. No subscription.
01 / OVERVIEW

Kimi K3
overview.

Kimi K3 is an open-weight reasoning model from Moonshot AI. Compared with Kimi K2.6, it has a larger published context window and a higher result on the Artificial Analysis index below. On AELDER, you can compare their current token rates and use either model with the same API key and prepaid balance.

Published model specifications

Context window1M tokens
ReasoningSupported
Inputtext & image
Outputtext
AA Intelligence Index44 / v4.3.2
Catalog release16 Jul 2026
Source: Artificial Analysis ↗
OPTIONAL / CONNECTIONAPI request example

Set the model ID and base URL in your OpenAI-compatible client.

BASE URL
https://aelder.dev/v1
API documentation ↗
import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["AELDER_API_KEY"],
    base_url="https://aelder.dev/v1",
)

response = client.chat.completions.create(
    model="kimi-k3",
    messages=[{"role": "user", "content": "Hello!"}],
)

print(response.choices[0].message.content)
MODELkimi-k3Chat Completions
02 / MODEL BENCHMARKS

Kimi K3 benchmarks.

Compare models in the AELDER catalog on the same external measure. The Artificial Analysis Intelligence Index combines evaluations of model capability. Higher is better; the score is not a percentage.

Artificial Analysis Intelligence Index

VERSION 4.3.2 · INDEX POINTS · HIGHER IS BETTER
Kimi K344
41 scored catalog models
Catalog models and their external Artificial Analysis evaluation configurations
ModelSource configurationIndex
Claude Opus 5.5Claude Opus 5.5 (Max, Default Fallback)58
Claude Sonnet 5.5Claude Sonnet 5.5 (Max, Default Fallback)56
Claude Fable 5.1Claude Fable 5.1 (Max, Default Fallback)53
GPT 6 AstraGPT-6 Astra (Max)53
GPT 6.1 SolGPT-6.1 Sol (Max)52
Claude Opus 5Claude Opus 5 (Max)51
Claude Fable 5Claude Fable 5 (Max, Opus 4.8 Fallback)50
GPT 6 SolGPT-6 Sol (Max)48
GPT 5.6 SolGPT-5.6 Sol (Max)47
DeepSeek V4.1 FlashDeepSeek V4.1 Flash 1017 (Max)45
GLM 5.3GLM-5.3 (Max)45
Qwen3.8 Max 0902Qwen3.8 Max (0902)45
Grok 4.6Grok 4.6 (Xhigh)44
Kimi K3CurrentKimi K3 (Max)44
Claude Haiku 5.5Claude Haiku 5.5 (Max)43
Claude Opus 4.8Claude Opus 4.8 (Max)42
GLM 5.3 FlashGLM-5.3-Flash42
GPT 5.6 TerraGPT-5.6 Terra (Max)42
Claude Opus 4.7 (1M)Claude Opus 4.7 (Max)41*
Gemini 3.8 FlashGemini 3.8 Flash (High)41
Qwen3.8 2.4T-A95BQwen3.8 2.4T A95B40
Qwen3.8 MaxQwen3.8 Max40
Claude Sonnet 5Claude Sonnet 5 (Max)38
GPT 6 LunaGPT-6 Luna (Max)38
GPT 5.6 LunaGPT-5.6 Luna (Max)37
DeepSeek V4 ProDeepSeek V4 Pro 0813 (Max)36
GLM 5.2GLM-5.2 (Max)34
Qwen3.8 27BQwen3.8 27B (Xhigh)34
GPT 5.3 CodexGPT-5.3 Codex (Xhigh)33*
Claude Opus 4.6Claude Opus 4.6 (Max)32*
Claude Sonnet 4.6Claude Sonnet 4.6 (Max)30
Gemini 3.1 ProGemini 3.1 Pro Preview30
Claude Opus 4.5Claude Opus 4.5 (Reasoning)29*
MiniMax M3MiniMax-M329
Kimi K2.6Kimi K2.6 (Reasoning)27
GLM 5.1GLM-5.1 (Reasoning)26
Kimi K2.7 CodeKimi K2.7 Code26
Hy3Hy325
DeepSeek V4 FlashDeepSeek V4 Flash 0420 (Max) · April 2026 reference24
Kimi K2.5Kimi K2.5 (Reasoning)23*
MiniMax M2.7MiniMax-M2.723
Current modelScroll to explore all models →Sources checked 21 Oct 2026

DeepSeek V4 Flash uses the April 0420 release reference, labelled in the chart and table. External results refer to the labelled source configurations, which may differ from AELDER’s configuration. They do not measure AELDER API speed or uptime. Source score markers (*) are preserved.

Why are 4 catalog models missing a score?

41 of 45 catalog models have a source reference on this index version. A missing result is not a zero score. These models remain in the catalog.

KEEP EXPLORING

Kimi K3 alternatives.

Explore all models ↗
MODEL DETAILS

Kimi K3
questions, answered.

How do Kimi K3 and Kimi K2.6 compare on price?
On AELDER, K3 input costs $0.75 per million tokens, compared with $0.19 for K2.6. Output costs $3.75 versus $0.80. K3 scores 44 on this external index; K2.6 scores 27, in their labelled evaluation configurations. Compare Kimi K2.6 API pricing alongside the chart to weigh the difference for your workload.
Is “Kimi K3 Max” a different model I need to select?
Use kimi-k3 as the model ID on AELDER. “Max” identifies the reasoning configuration used for the external benchmark. The label does not mean every API request runs with that same setting.
Does Kimi K3 generate images?
No. Its published output is text. Image understanding means interpreting visual input and returning a text response, rather than creating an image.
BUILD WITH AELDER

Your next idea,
one API away.

Get API key

One key. Multiple models. Pay as you go.