Pipe API docs — OpenAI-compatible chat

OpenAI-compatible inference on Pipe. Base URL https://api.pipenetwork.ai/v1. Sign in with Google, create a key at /keys, and POST /v1/chat/completions. Default model is pipe/deepseek-v4-flash-0731 (tools on, no vision). POST /v1/tasks submits a code.change outcome contract. Project tickets on the dashboard Board dispatch to hi tickets or sandbox /v1/tasks. Home GPU providers: Cloudflare Tunnel terminates TLS in front of Ollama or llama.cpp; list the public HTTPS URL on /providers.

First call

  1. Sign in with Google
  2. Create a key at /keys
  3. POST /v1/chat/completions with model pipe/deepseek-v4-flash-0731

Base URL

https://api.pipenetwork.ai/v1

Authentication

Authorization: Bearer $PIPE_API_KEY

Quickstart

curl https://api.pipenetwork.ai/v1/chat/completions \
  -H "Authorization: Bearer $PIPE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "pipe/deepseek-v4-flash-0731", "messages": [{"role": "user", "content": "Hello"}]}'

OpenAI SDK

from openai import OpenAI
client = OpenAI(base_url="https://api.pipenetwork.ai/v1", api_key=os.environ["PIPE_API_KEY"])
resp = client.chat.completions.create(model="pipe/deepseek-v4-flash-0731", messages=[{"role": "user", "content": "Hello"}])

Home · Models & pricing · x402 · Fine-tuning · Verified tasks · Project tickets · Provider guide · Commerce guide