NewTimeSpace API Interface Introduction

Quick Start (AI Agent / Developer Edition)

Base URL: https://www.newtimespace.comNo registration, no API key; GET requests only.

First call, with explicit parameters:

GET https://www.newtimespace.com/api/v1/article/getArticles?page=1&limit=10&language=en&format=json

Up and running in ten seconds (Python):

import json, urllib.request

raw = urllib.request.urlopen(
        "https://www.newtimespace.com/api/v1/article/getArticles?format=json"
).read()
data = json.loads(raw.decode("utf-8-sig")) # responses start with a UTF-8 BOM; use utf-8-sig
for a in data["articles"]:
        print(a["article_id"], a["pubDate"], a["title"])
print(data["pagination"]) # use the pages field as the pagination stop condition

Note: every response (JSON, XML, success and error alike) starts with a UTF-8 BOM (EF BB BF). Strip it before parsing, or JSON parsing fails with "Expecting value".

Recommended call flow

Note: list results already contain the full article text (content is the complete HTML). "Search → read the text" normally needs the list endpoint only; call the details endpoint only to fetch by known ID, to batch-fetch (up to 20 at a time), or to switch language versions.

Two endpoints — which one to use?

Scenario 1: Discovering news

Search English articles whose title contains "profit warning" (query values must be URL-encoded):

GET /api/v1/article/getArticles?title=profit%20warning&language=en&format=json

Fetch ETF articles by category (category must match the language — see the full documentation):

GET /api/v1/article/getArticles?category=ETF&language=en&format=json

English articles published on a given day, 20 per page:

GET /api/v1/article/getArticles?language=en&date=2026-09-08&limit=20&format=json

Response structure: status + an articles array + a pagination object. Results are always sorted by publish time in descending order. No matches return HTTP 200 with an empty array — a normal response; do not retry.

Scenario 2: Fetching a specific article

Take an article_id from list results (e.g. 1445593) and fetch by ID:

GET /api/v1/article/detail?articleIds=1445593&format=json

Returns status + data, where data is the article object (title, description, content, pubDate, url and other full fields). Note: article_id must come from real list results — never construct one; a nonexistent ID returns HTTP 200 with "data": []; treat it as "not found".

Scenario 3: Fetching multiple articles in one request

Separate multiple IDs with commas:

GET /api/v1/article/detail?articleIds=1445593,1445506&format=json

Up to 20 IDs per request; more returns 400 (articleIds cannot exceed 20 items). Split larger batches, e.g. 45 IDs into 20 + 20 + 5.

Note: when a batch matches exactly 1 article, data is an object rather than an array; invalid IDs in a batch are skipped and only hits are returned — the shape of data follows the hit count. Check the data type before parsing.

Notes for agents

• Responses carry a UTF-8 BOM: strip it before parsing; prefer a decoding mode that removes the BOM automatically (e.g. Python's utf-8-sig).
• Time fields are Beijing time (GMT+8): pubDate / date / time are all Beijing time; the pubDateTZ field is currently fixed to "UTC" — do not convert based on it, and do not add 8 hours.
• The response format is determined solely by the format parameter: always pass format=json explicitly; the Accept request header does not affect the format.
• Empty results are normal: no matches, a category/language mismatch, or a page beyond the total all return HTTP 200 with an empty array — do not treat them as failures or retry; stop paging on pagination.pages instead of reading only the first page.
• Detail data has two shapes: an object when exactly 1 article matches, an array when several match; parsing code must handle both.
• Rate limit: 20 requests per IP per 60 seconds; exceeding it returns 429 with a Retry-After header (wait accordingly before retrying). The response carries no remaining-quota headers — count request frequency yourself.
• GET only: a POST request returns HTTP 200 plain text illegal request., which is not JSON.