Hormouz AIHormouzSign in

MoonshotAI: Kimi K2.5: AI model

Kimi K2.5 is a native multimodal model strong at visual coding and agent swarms from Moonshot AI. Reachable from inside Iran through Hormouz.

Context window:
262,144 tokens
Input:
$0.5/1M tokens
Output:
$2.5/1M tokens
Type:
text, image
  • Vision
  • Tools
  • Streaming
  • Long context

Thinking levels:

Kimi K2.5 is a natively multimodal model from Moonshot AI. It extends Kimi K2 through continued pretraining on roughly 15 trillion tokens of mixed image and text data, and pairs leading visual-coding ability with a self-directed agent-swarm approach. Over a 256K-token context it keeps strong general reasoning and agentic tool-calling, and it takes both text and image input.

Kimi K2.5 is a good fit for visual coding, multimodal agents and UI-to-code workflows. It supports tool (function) calling, image input and explicit reasoning. The model works with a context window of about 262K tokens, so it comfortably handles long inputs. It is served through Hormouz behind an OpenAI-compatible API. Just point your base URL at Hormouz and drop in a Hormouz key, and existing OpenAI SDK code runs unchanged.

Pricing is usage-based: $0.6 per million input tokens and $3.5 per million output tokens, billed only for what you use. Hormouz provides stable access to the model from inside Iran, with no foreign credit card or circumvention tooling required.

Kimi K2.5 is Moonshot AI's native multimodal model, delivering state-of-the-art visual coding capability and a self-directed agent swarm paradigm. Built on Kimi K2 with continued pretraining over approximately 15T mixed...

Quickstart

curl https://api.hormouz.net/v1/chat/completions \
  -H "Authorization: Bearer ng-YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "moonshotai/kimi-k2.5",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

Specs

VendorMoonshot AI
TypeText, Image
Context window262,144 tokens
Pricing$0.50/M in · $2.50/M out
APIChat Completions
CapabilitiesStreaming · Vision · Tools

Benchmarks

46.8
Coding Index
Better than 51% of models we compared

Reasoning

GPQA Diamond
Graduate-level scientific reasoning
87.9%
HLE
Humanity's Last Exam
30.7%
IFBench
Instruction-following
70.2%
AA-LCR
Long-context reasoning
78%
CritPt
Research-level physics reasoning
3.1%
AA Omniscience
Broad factual knowledge accuracy
35.2%
Non-Hallucination
Rate of avoiding fabricated answers
34.3%

Coding

Terminal-Bench Hard
Agentic coding & terminal use
34.8%

Agentic

τ²-Bench
Tool-use agents in dual-control scenarios
95.9%
GDPval
Economically valuable tasks
17.2%

Source: Artificial Analysis

Ranking distribution

First
9,724
Second
9,169
Third
5,082
Fourth
3,257

27232 tournaments ?

Android Native
Top 70%
1,133Elo
Win rate: 57.8%
Full-Stack
Top 66%
1,112Elo
Win rate: 54.5%
Godot Game Dev
Top 42%
1,219Elo
Win rate: 59.8%
Mobile Apps
Top 63%
1,144Elo
Win rate: 54.7%
Web Apps
Top 69%
1,129Elo
Win rate: 51.6%
3D
Top 32%
1,216Elo
Win rate: 53.1%Avg time: 148s
ASCII Art
Top 43%
1,177Elo
Win rate: 46.4%Avg time: 285s
Code
Top 27%
1,242Elo
Win rate: 54.1%Avg time: 218s
Data Viz
Top 33%
1,228Elo
Win rate: 51.3%Avg time: 187s
Game Dev
Top 31%
1,219Elo
Win rate: 53.4%Avg time: 229s
SVG
Top 43%
1,149Elo
Win rate: 48.4%Avg time: 107s
UI Component
Top 30%
1,243Elo
Win rate: 53.6%Avg time: 192s
Website
Top 24%
1,255Elo
Win rate: 55.2%Avg time: 246s

Metrics sourced from Design Arena

Code example

cURL
curl https://api.hormouz.net/v1/chat/completions \
  -H "Authorization: Bearer ng-YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "moonshotai/kimi-k2.5",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'
Python
from openai import OpenAI
client = OpenAI(api_key="ng-YOUR_API_KEY", base_url="https://api.hormouz.net/v1")
response = client.chat.completions.create(
    model="moonshotai/kimi-k2.5",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(response.choices[0].message.content)
JavaScript
import OpenAI from "openai";
const client = new OpenAI({ apiKey: "ng-YOUR_API_KEY", baseURL: "https://api.hormouz.net/v1" });
const response = await client.chat.completions.create({
  model: "moonshotai/kimi-k2.5",
  messages: [{ role: "user", content: "Hello!" }],
});
console.log(response.choices[0].message.content);

Buy the MoonshotAI: Kimi K2.5 API, paid in rial

  • Buy credit with an Iranian bank card; no foreign card or account needed.
  • One Hormouz key reaches MoonshotAI: Kimi K2.5 and 600+ other models, and you pay only for what you use.
  • No VPN and no risk of a banned account; just change the base URL to api.hormouz.net/v1.
Create an account

Use via Hormouz

Connect to this model and 600+ others with one API key, then see usage down to each request.

Request company access

FAQ

How much does Kimi K2.5 cost?
Billing is usage-based: $0.6 per million input tokens and $3.5 per million output tokens. There is no fixed monthly fee. You pay only for what you use.
How do I access Kimi K2.5 from inside Iran?
Create a Hormouz account and generate an API key. Hormouz provides a stable path to the model from inside Iran, with no foreign credit card or circumvention tooling required.
Is Kimi K2.5 compatible with OpenAI code?
Yes. Hormouz exposes an OpenAI-compatible API. Change the base URL to Hormouz and swap in your Hormouz key. Your existing SDKs and code run without a rewrite.
What context window and capabilities does Kimi K2.5 have?
Kimi K2.5 works with a context window of about 262K tokens. It also accepts image input, supports tool calling, can produce explicit reasoning. That makes it suitable for visual coding, multimodal agents and UI-to-code workflows.
What is Kimi K2.5 good for?
It is best suited to visual coding, multimodal agents and UI-to-code workflows.
Which provider serves Kimi K2.5 on Hormouz?
Hormouz is the provider you connect to and bill through. It routes your request to the model and handles access, billing and stable connectivity from inside Iran on your behalf.

Updated 2026-10-13