Getting A-share quotes with Python: three approaches

Option 1: call the HTTP API directly (recommended)

Free endpoints need no key: one GET request returns price, OHLC, volume, turnover value and turnover rate. The response is one envelope `{ ok, endpoint, tier, elapsed_ms, source, data }` where `data` is an array sorted newest-first, so `data[0]` is the latest (current session) bar.

The code is a few lines (free endpoints need no auth header):

import requests

r = requests.get(
    "https://api.ashareapi.com/v1/quote",
    params={"code": "sh600667"},   # market prefix: sh / sz / bj / hk / us
    timeout=10,
)
body = r.json()
if not body.get("ok"):
    raise RuntimeError(body)       # upstream failure returns ok=false and is NOT counted

bar = body["data"][0]              # latest (current session) bar
print(bar["date"], bar["last"], bar["turnover"])

• `last` = latest price · `open/high/low` = OHLC · `turnover` = turnover rate (%)
• `volume` is in lots (×100 = shares) · `amount` is in CNY
• `date` carries the data date — on a market holiday it will not say "today"; do not treat the previous session as live

Option 2: use an existing data SDK (Tushare / AkShare)

If you already use an SDK there is no rush to switch — its strength is breadth (minute bars, financial statements, news). The cost is two other things:

Tushare requires signup plus credit management, and real-time data is a separate paid permission; AkShare needs no token, but it scrapes public pages, so upstream redesigns break functions and you maintain them (that is most of its issue tracker).

# pip install akshare
import akshare as ak

df = ak.stock_zh_a_spot_em()        # whole-market snapshot, filter yourself
row = df[df["代码"] == "600667"]
print(row)

• When not to use an SDK: if you only need mainstream data (quotes / K-line / money flow / top-trader boards) and do not want to maintain a pipeline — calling an API is less work
• When you must use one: you need minute bars or full news / filings / research text — we do not offer those (see the FAQ boundary below)

Option 3: scrape quote pages yourself (not recommended unless there is no API)

Scraping loses three ways when an API exists: page structure changes without notice, anti-bot measures block you (rate limits, captchas, IP bans), and the compliance call for how you use the data lands back on you.

It is only worth it when the data genuinely has no API (niche instruments, odd historical snapshots). Even then, look for an official data service first instead of starting from a page.

How to choose

Common errors and fixes

• `429` = rate limited: anonymous 5/min; solve one PoW challenge (`GET /v1/challenge`) for 60/min; keys follow their tier (from 10k calls)
• `401` = you called a key-required endpoint without a key (quote / K-line / hot list / market overview / breadth need none)
• Codes need a market prefix: `sh600667` / `sz000001` / `bj830799` / `hk00700` / `usAAPL`; a bare 6-digit code will not resolve
• `ok:false` = upstream fetch failed: we already failed over and it is not counted; one retry usually succeeds
• Holidays and intraday: the `date` field tells you which session the data belongs to — do not build logic on "today"