START IN SECONDS.

AGENT SETUP

Read and set up feihoa.com/llms.txt

Paste into your agent for automatic setup.

Doing it yourself? Good. Start here.

Connect to Qwen3.8 27B FP8 Uncensored through an OpenAI-compatible endpoint built for coding, tool use, and long agent runs. Streaming is built in. A monthly usage meter is not.

BASE URL
MODEL
LIMITS: Shared across your account and included in each API response
RETRIES: Run requests sequentially; honor Retry-After and reuse the same Idempotency-Key when retrying
LONG RUNS: Use streaming and process chunks as they arrive
OPENCODE: Set timeout to false, chunkTimeout to 30000, and model limits from /v1/models

VERIFY WITH CURL

curl https://api.feihoa.com/v1/chat/completions \
  -H "Authorization: Bearer $FEIHOA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Qwen3.8-27B-Uncensored",
    "messages": [{"role": "user", "content": "Reply with FEIHOA connected."}],
    "stream": true
  }'