Guide
Kimi K2.5 API: Moonshot's Agentic AI Model (256K Context)
April 16, 2026 · 4 min read
Kimi K2.5 from Moonshot AI is designed for agentic tasks — multi-step reasoning, tool use, and complex workflows. With a 256K context window and competitive pricing, it's a strong choice for AI agent builders.
Kimi K2.5 Specs
| Feature | Value |
|---|---|
| Context Window | 256K tokens |
| Input Price | $0.24/M tokens |
| Output Price | $1.20/M tokens |
| Best For | Agentic tasks, multi-step reasoning |
| Provider | Moonshot AI (China) |
Quick Start
from openai import OpenAI
client = OpenAI(base_url="https://api.aipower.me/v1", api_key="YOUR_KEY")
response = client.chat.completions.create(
model="moonshot/kimi-k2.5",
messages=[
{"role": "system", "content": "You are an AI agent. Break tasks into steps."},
{"role": "user", "content": "Research and summarize the latest AI news"},
],
)
print(response.choices[0].message.content)Kimi vs Other Agentic Models
| Model | Context | Price (in/out) | Agentic Score |
|---|---|---|---|
| Kimi K2.5 | 256K | $0.24/$1.20 | Strong |
| Claude Opus | 200K | $7.50/$37.50 | Best |
| GPT-5.4 | 272K | $3.75/$22.50 | Strong |
Kimi K2.5 offers 80-97% cost savings vs Claude Opus/GPT-5.4 for agentic workflows.
Try it with 50 free API calls at aipower.me.