> For the complete documentation index, see [llms.txt](https://docs.usedecentral.org/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.usedecentral.org/integrations/python.md). # Python Decentral works with the standard OpenAI Python SDK. For additional functionality like structured receipt access and async streaming, you can also use `httpx` directly. *** ## Using the OpenAI Python SDK ```bash pip install openai ``` ```python import os from openai import OpenAI client = OpenAI( base_url="https://api.usedecentral.org/v1", api_key=os.environ["DECENTRAL_API_KEY"] ) # Non-streaming response = client.chat.completions.create( model="qwen3-8b", messages=[ {"role": "system", "content": "You are a precise technical assistant."}, {"role": "user", "content": "Explain Solana accounts vs Ethereum contracts."} ], max_tokens=512, temperature=0.3 ) print(response.choices[0].message.content) print(f"Tokens used: {response.usage.total_tokens}") ``` ### Streaming ```python stream = client.chat.completions.create( model="llama-3.3-70b", messages=[{"role": "user", "content": "Write a Python function to parse JSON safely."}], stream=True ) for chunk in stream: delta = chunk.choices[0].delta.content if delta: print(delta, end="", flush=True) print() ``` ### Async ```python import asyncio from openai import AsyncOpenAI async def main(): client = AsyncOpenAI( base_url="https://api.usedecentral.org/v1", api_key=os.environ["DECENTRAL_API_KEY"] ) async with client.chat.completions.stream( model="qwen3-8b", messages=[{"role": "user", "content": "What is a Merkle tree?"}] ) as stream: async for chunk in stream: delta = chunk.choices[0].delta.content if delta: print(delta, end="", flush=True) asyncio.run(main()) ``` *** ## Accessing on-chain headers with httpx The OpenAI SDK does not expose response headers. Use `httpx` directly when you need the Solana transaction signatures. ```python import os import httpx API_KEY = os.environ["DECENTRAL_API_KEY"] BASE_URL = "https://api.usedecentral.org/v1" def chat(messages: list, model: str = "qwen3-8b") -> dict: response = httpx.post( f"{BASE_URL}/chat/completions", headers={ "Authorization": f"Bearer {API_KEY}", "Content-Type": "application/json" }, json={"model": model, "messages": messages}, timeout=120 ) response.raise_for_status() data = response.json() data["_decentral"] = { "job_id": response.headers.get("x-decentral-job-id"), "escrow_tx": response.headers.get("x-decentral-tx-signature"), "settlement_tx": response.headers.get("x-decentral-settlement-tx"), "worker": response.headers.get("x-decentral-worker"), "credits_remaining": response.headers.get("x-decentral-credits-remaining") } return data result = chat([{"role": "user", "content": "What is Solana PoH?"}]) content = result["choices"][0]["message"]["content"] job_id = result["_decentral"]["job_id"] settlement = result["_decentral"]["settlement_tx"] print(content) print(f"\nVerify payment: https://explorer.solana.com/tx/{settlement}") ``` *** ## Streaming with httpx ```python import json import httpx def stream_chat(messages: list, model: str = "qwen3-8b"): with httpx.stream( "POST", f"{BASE_URL}/chat/completions", headers={ "Authorization": f"Bearer {API_KEY}", "Content-Type": "application/json" }, json={"model": model, "messages": messages, "stream": True}, timeout=120 ) as response: response.raise_for_status() for line in response.iter_lines(): if not line or line == "data: [DONE]": continue if line.startswith("data: "): chunk = json.loads(line[6:]) delta = chunk["choices"][0]["delta"].get("content", "") if delta: print(delta, end="", flush=True) stream_chat([{"role": "user", "content": "What is WebGPU?"}]) ``` *** ## Multi-turn conversations ```python from openai import OpenAI client = OpenAI( base_url="https://api.usedecentral.org/v1", api_key=os.environ["DECENTRAL_API_KEY"] ) messages = [{"role": "system", "content": "You are a knowledgeable assistant."}] while True: user_input = input("You: ").strip() if not user_input: break messages.append({"role": "user", "content": user_input}) response = client.chat.completions.create( model="qwen3-8b", messages=messages ) assistant_message = response.choices[0].message.content messages.append({"role": "assistant", "content": assistant_message}) print(f"Assistant: {assistant_message}") ``` *** ## Error handling ```python from openai import OpenAI, APIStatusError, APIConnectionError import time client = OpenAI( base_url="https://api.usedecentral.org/v1", api_key=os.environ["DECENTRAL_API_KEY"] ) def chat_with_retry(messages: list, model: str, max_retries: int = 3) -> str: for attempt in range(max_retries): try: response = client.chat.completions.create( model=model, messages=messages ) return response.choices[0].message.content except APIStatusError as e: if e.status_code == 402: raise Exception("Insufficient credits. Top up at usedecentral.org/app") from e if e.status_code == 503: retry_after = e.response.json().get("error", {}).get("details", {}).get("retry_after", 10) if attempt < max_retries - 1: time.sleep(retry_after) continue raise if e.status_code == 504: # Job timed out, credits were refunded if attempt < max_retries - 1: time.sleep(5) continue raise raise except APIConnectionError as e: if attempt < max_retries - 1: time.sleep(2 ** attempt) continue raise raise Exception(f"Failed after {max_retries} attempts") ``` *** ## Checking account balance ```python import httpx def get_balance() -> dict: response = httpx.get( "https://api.usedecentral.org/v1/account", headers={"Authorization": f"Bearer {os.environ['DECENTRAL_API_KEY']}"} ) response.raise_for_status() return response.json() balance = get_balance() print(f"Credits remaining: {balance['credits_remaining']} (${balance['usdc_value']:.2f})") ``` --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://docs.usedecentral.org/integrations/python.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.