OpenAI ប្រសិទ្ធភាពខ្ពស់៖ គំរូ GPT-5.4 Mini (ទស្សន៍ទ្រនិច) ពី OpenAI ដែលមានបង្អួចបរិបទនិមិត្តសញ្ញា 400,000 ជាមួយនឹងភាពឆបគ្នានៃ OpenAI API ពេញលេញ។ (វ៉ារ្យ៉ង់ដែលបានធ្វើឱ្យប្រសើរជាពិសេសសម្រាប់ Cursor IDE និងលំហូរការងារភ្នាក់ងារ) ។
Context Window
400K tokens
Max Output Tokens
16,384 tokens
Release Date
2025 - 2026
Step-by-step Reasoning
Standard
• Frontier Class• Top Tier
$ Token Pricing (Per 1M tokens)
Pay-As-You-Go pricing based on actual usage. No recurring maintenance fees, no minimum commitment, automatic real-time wallet deduction.
Input (Prompt Tokens)
2.019 ₫/M$0.0783/M
6.250 ₫/M$0.25/M69% OFF
Output (Completion Tokens)
2.019 ₫/M$0.0783/M
25.000 ₫/M$1.00/M92% OFF
Official provider list price: $0.25 / $1.00 per 1M tokens. Save up to 69% via Gateway.
24h Gateway Performance
Real-world latency and throughput metrics automatically monitored 24/7.
Average TPS
97.3 t/s
Average Latency
12.94 s
Time to First Token
6.28 s
Success Rate
100.00%
Quickstart API Integration
Compatible with standard OpenAI SDKs. Replace the Base URL and pass Model ID gpt-5.4-mini-cursor:
curl https://giahuynexus.com/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_API_KEY" \
-d '{
"model": "gpt-5.4-mini-cursor",
"messages": [
{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": "Hello, how can you help me today?"}
],
"temperature": 0.7
}'
from openai import OpenAI
client = OpenAI(
api_key="YOUR_API_KEY",
base_url="https://giahuynexus.com/v1"
)
response = client.chat.completions.create(
model="gpt-5.4-mini-cursor",
messages=[
{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": "Hello, how can you help me today?"}
],
temperature=0.7
)
print(response.choices[0].message.content)
import OpenAI from "openai";
const openai = new OpenAI({
apiKey: "YOUR_API_KEY",
baseURL: "https://giahuynexus.com/v1",
});
const response = await openai.chat.completions.create({
model: "gpt-5.4-mini-cursor",
messages: [
{ role: "system", content: "You are a helpful assistant." },
{ role: "user", content: "Hello, how can you help me today?" }
],
temperature: 0.7,
});
console.log(response.choices[0].message.content);
OpenAI: GPT-5.4 Mini (Cursor)
gpt-5.4-mini-cursor
Copy the cURL command below and run it in your terminal to test model responses: