Cupang Proxy
OpenAI-compatible LLM gateway. 20 native tool-calling models, ready to use.
⬦ created by Satowi Nakamoto
20Models
OpenAICompatible API

Endpoint

POST https://cupangproxy.tailfab3a7.ts.net/v1/chat/completions

Auth: Authorization: Bearer YOUR_API_KEY

🔑 Request an API key from the admin. Public keys are rate-limited (30 req/min, 2K req/day by default).

Available Models (20)

ModelLatency
cp/deepseek-v4-flash <5s
cp/deepseek-v4-pro <5s
cp/gemma-4-31b 5-15s
cp/glm-5 15-45s
cp/glm-5.1 5-15s
cp/glm-5.2 5-15s
cp/hermes-4-70b 15-45s
cp/kimi-k2-thinking <5s
cp/kimi-k2.5 5-15s
cp/kimi-k2.6 5-15s
cp/kimi-k2.7-code 5-15s
cp/minimax-m2.5 15-45s
cp/minimax-m2.7 5-15s
cp/minimax-m3 15-45s
cp/nemotron-3-super-120b 15-45s
cp/qwen3-coder-30b <5s
cp/qwen3-coder-480b <5s
cp/qwen3-coder-next <5s
cp/qwen3.5-397b 5-15s
cp/qwen3.6-35b 5-15s

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/deepseek-v4-flash",
    "messages": [{"role":"user","content":"Hello!"}]
  }'

Rate Limits