GPTOSS-Proxy OpenAI-Compatible

Leave a ⭐ if this helped you

A tiny Cloudflare Worker that exposes a single OpenAI-compatible endpoint:

• POST /v1/chat/completions → supports stream: true|false, token streaming, and reasoning (CoT) as reasoning_content.
• GET /v1/models → lists available models: gpt-oss-120b, gpt-oss-20b.
• GET / → returns { success: true, discord, website, repo }.

🚀 One-click deploy

NOTE: You can Join our Discord server if you face issues deploying.

✨ Features

• OpenAI-compatible /v1/chat/completions (streaming & non-stream)
• CoT reasoning stream → choices[0].delta.reasoning_content
• Models: gpt-oss-120b (larger) and gpt-oss-20b (faster)
• Reasoning effort control: none | low | medium | high
• Thread persistence via headers or payload metadata
• CORS enabled for browser clients

▶️ Endpoints

GET /

Returns:

{
  "success": true,
  "discord": "https://discord.gg/cwDTVKyKJz",
  "website": "https://ish.junioralive.in",
  "repo": "https://github.com/junioralive/gptoss-proxy"
}

GET /v1/models

Lists supported models:

{ "object":"list", "data":[
  {"id":"gpt-oss-120b","object":"model"},
  {"id":"gpt-oss-20b","object":"model"}
]}

POST /v1/chat/completions

OpenAI-style body:

{
  "model": "gpt-oss-20b",
  "messages": [
    {"role": "user", "content": "Say hi and then a space fact."}
  ],
  "stream": true,
  "metadata": {
    "gptoss_user_id": "u123",         // optional
    "gptoss_thread_id": "thr_456",    // optional
    "reasoning_effort": "medium"      // none|low|medium|high
  }
}

Optional headers (override metadata):

• X-GPTOSS-User-Id: <uuid>
• X-GPTOSS-Thread-Id: thr_...
• X-Reasoning-Effort: none|low|medium|high

Streaming response: OpenAI-style SSE

• token deltas → choices[0].delta.content
• reasoning updates → choices[0].delta.reasoning_content
• terminates with a chunk where finish_reason = "stop" then data: [DONE]

Non-streaming: standard chat.completion JSON.

🧪 Quick start (cURL)

Streaming:

curl -N https://<your-worker>.workers.dev/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "X-Reasoning-Effort: medium" \
  -d '{"model":"gpt-oss-120b","messages":[{"role":"user","content":"hello!"}],"stream":true}'

Non-stream:

curl https://<your-worker>.workers.dev/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-oss-20b","messages":[{"role":"user","content":"One short sentence."}],"stream":false}'

🐍 Python example

import os, json, requests

BASE = "https://<your-worker>.workers.dev"

# stream
payload = {
    "model": "gpt-oss-120b",
    "messages": [{"role":"user","content":"Say hi, then a fun space fact."}],
    "stream": True,
    "metadata": {"reasoning_effort": "medium"}
}
with requests.post(f"{BASE}/v1/chat/completions",
                   headers={"Content-Type":"application/json"},
                   data=json.dumps(payload), stream=True, timeout=120) as r:
    r.raise_for_status()
    for line in r.iter_lines(decode_unicode=True):
        if not line: continue
        if line.startswith("data: "):
            data = line[6:].strip()
            if data == "[DONE]":
                print("\n[DONE]"); break
            obj = json.loads(data)
            delta = obj["choices"][0]["delta"]
            if "reasoning_content" in delta:
                print("\n[reasoning]", delta["reasoning_content"])
            if "content" in delta:
                print(delta["content"], end="", flush=True)

# non-stream
payload = {"model":"gpt-oss-20b","messages":[{"role":"user","content":"Now reply in one short sentence."}],"stream":False}
print("\n\nNon-stream result:")
print(requests.post(f"{BASE}/v1/chat/completions",
                    headers={"Content-Type":"application/json"},
                    json=payload, timeout=60).json())

🤝 Use with OpenAI SDKs

Point the SDK to your Worker:

Python

from openai import OpenAI
client = OpenAI(base_url="https://<your-worker>.workers.dev/v1", api_key="dummy")
stream = client.chat.completions.create(
  model="gpt-oss-20b",
  messages=[{"role":"user","content":"hello!"}],
  stream=True,
  metadata={"reasoning_effort": "low"}
)
for chunk in stream:
    d = chunk.choices[0].delta
    if getattr(d, "reasoning_content", None): print("\n[reasoning]", d.reasoning_content)
    if getattr(d, "content", None): print(d.content, end="")

Node

import OpenAI from "openai";
const client = new OpenAI({ baseURL: "https://<your-worker>.workers.dev/v1", apiKey: "dummy" });
const stream = await client.chat.completions.create({
  model: "gpt-oss-120b",
  messages: [{ role: "user", content: "hello!" }],
  stream: true,
  metadata: { reasoning_effort: "medium" }
});
for await (const chunk of stream) {
  const d = chunk.choices[0].delta;
  if (d.reasoning_content) console.log("\n[reasoning]", d.reasoning_content);
  if (d.content) process.stdout.write(d.content);
}

⚠️ Disclaimer

This script is provided as a proof of concept. It is not intended to cause harm or be misused.

If you encounter any issues or have concerns, please contact: 📧 [email protected] 💬 Join our Discord

🔧 Notes & tips

• Models: choose gpt-oss-20b for speed; gpt-oss-120b for quality.
• Reasoning effort: set via header or metadata.reasoning_effort (default medium).
• Threads / Users: pass X-GPTOSS-Thread-Id + X-GPTOSS-User-Id (or metadata.gptoss_*) to keep conversation history.
• CORS: enabled by default. If you add auth, include authorization in access-control-allow-headers.