Cupang Proxy — Curated LLM Gateway

Cupang Proxy

Curated OpenAI-compatible gateway. 14 live models, key-pool powered, latency tracked.

⚡ OpenAI-compatible · BYO key

14Models

2451Total Requests

133.0MTokens Served

3In Flight

⚡ Endpoint

POSThttps://cupangproxy.tailfab3a7.ts.net/v1/chat/completions

Auth: bring your own key — Authorization: Bearer YOUR_API_KEY

Recommended models tap to copy

Available Models (14)

waiting for live data…

Latency by model — fastest first

Model	Live Latency warming
`cp/deepseek-v3.2`	warming
`cp/deepseek-v4-flash`	warming
`cp/deepseek-v4-pro`	47.8s
`cp/glm-4.7-flash`	4.3s
`cp/glm-5.1`	5.1s
`cp/glm-5.2`	105.8s
`cp/gpt-oss-120b`	3.4s
`cp/kimi-k2.5`	warming
`cp/kimi-k2.6`	warming
`cp/kimi-k2.7-code`	12.1s
`cp/minimax-m2.7`	warming
`cp/minimax-m3`	warming
`cp/nemotron-3-super-120b`	14.4s
`cp/step-3.5-flash`	8.8s

Example (curl)

curl -X POST https://cupangproxy.tailfab3a7.ts.net/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "cp/glm-4.7-flash",
    "messages": [{"role":"user","content":"Hello!"}]
  }'

Client Config

{
  "base_url": "https://cupangproxy.tailfab3a7.ts.net/v1",
  "api_key": "YOUR_API_KEY",
  "model": "cp/glm-4.7-flash"
}