How to Automate Supply Chain Risk Reports: A Guide for Developers
Do you use Python? If so, this guide will help you automate supply chain risk reports using AI Chat GPT and our News API.
GNews API is a straightforward option for developers building a basic feed of mainstream headlines. Webz.io is built for a harder problem: finding news across a much larger source base, then giving developers the data and controls needed to turn that coverage into a production news feed, monitoring system, risk product, or AI application.
We tested both APIs across 83 live requests. In the matched coverage tests, Webz.io returned:
Webz.io also returned full extracted article text, detailed source metadata, entities, sentiment, categories, topics, stock tickers, trust signals, and syndication data.
The result is a clear distinction: GNews can supply a simple news feed. Webz.io supplies the news data layer for products that cannot afford large coverage gaps.
| Benchmark | Webz.io | GNews API |
| Title-only results across eight topics | 21,328 | 951 |
| Title and article-content results | 163,699 | 10,718 |
| Title-only coverage advantage | 22.4× | — |
| Broad-search coverage advantage | 15.3× | — |
| Topics where the provider returned more results | 8 of 8 | 0 of 8 |
| Hebrew “artificial intelligence” results | 414 | 7 |
| Spanish “artificial intelligence” results | 28,988 | 1,583 |
| Newest article with the tested credentials | 2 minutes 19 seconds old | 12 hours 14 minutes old |
| Median visible article-content length in the broad-search sample | 4,805 characters | 265 characters |
The freshness and content-length results reflect the credentials used in the test. The supplied GNews key was on its free plan, which has a documented 12-hour delay and automatically truncates article content. Paid GNews plans offer real-time availability and full content.
Read the full benchmark methodology and results
The coverage benchmark used a fixed 72-hour publication window:
August 31, 2026 at 00:00 UTC through September 2, 2026 at 23:59 UTC
The window ended more than 12 hours before the test was run. This prevented the GNews free plan’s delay from reducing its coverage counts.
We tested eight topics in English. Each topic was searched in two ways:
| Search topic | Webz.io | GNews | Webz.io advantage |
| Artificial intelligence | 846 | 21 | 40.3× |
| Climate change | 685 | 82 | 8.4× |
| Bitcoin | 5,243 | 143 | 36.7× |
| NVIDIA | 1,951 | 155 | 12.6× |
| Israel | 5,477 | 241 | 22.7× |
| Ukraine | 6,811 | 298 | 22.9× |
| Quantum computing | 198 | 7 | 28.3× |
| Rare disease | 117 | 4 | 29.3× |
| Total | 21,328 | 951 | 22.4× |
Title-only search is the fairest coverage comparison because it reduces differences in full-text extraction and article parsing. Webz.io still returned more articles for every topic, with advantages ranging from 8.4× to 40.3×.
| Search topic | Webz.io | GNews | Webz.io advantage |
| Artificial intelligence | 42,197 | 3,780 | 11.2× |
| Climate change | 21,559 | 1,606 | 13.4× |
| Bitcoin | 30,768 | 396 | 77.7× |
| NVIDIA | 12,818 | 664 | 19.3× |
| Israel | 22,296 | 1,801 | 12.4× |
| Ukraine | 28,026 | 2,158 | 13.0× |
| Quantum computing | 2,078 | 55 | 37.8× |
| Rare disease | 3,957 | 258 | 15.3× |
| Total | 163,699 | 10,718 | 15.3× |
The broad-search test shows the size of the difference when developers need to find mentions inside the body of an article, not only in its headline.
This matters for media monitoring, adverse-media screening, financial intelligence, brand monitoring, market research, and risk products. A local article may mention a company, executive, product, lawsuit, investigation, or disruption without putting that subject in the headline.
Webz.io returned more matches in all 16 paired coverage tests.
That result is consistent with the providers’ stated source footprints. Webz.io says its News API covers more than 350,000 news sources, compared with more than 80,000 sources stated by GNews. The providers may count sources differently, so those figures were not used to calculate the benchmark result. The live query totals provide the stronger comparison.
The difference is not limited to obscure queries. Webz.io returned:
A news API that covers fewer publishers may still return a useful set of major stories. But it cannot recover relevant articles that were never present in its index.
For a homepage, ten major headlines may be enough. For a monitoring or intelligence product, missed coverage becomes a product blind spot.
Webz.io states that its News API covers more than 170 languages across 200+ countries. GNews states coverage across 41 languages and 71 countries.
The live multilingual probes showed how that difference can affect actual searches:
| Query | Webz.io | GNews | Webz.io advantage |
| Hebrew phrase for “artificial intelligence” | 414 | 7 | 59.1× |
| Spanish phrase for “artificial intelligence” | 28,988 | 1,583 | 18.3× |
Two phrases are not a complete language benchmark. They do, however, show the difference between nominal language support and the amount of searchable news available in those languages.
This is especially relevant for products monitoring:
A global product needs more than translations of major international stories. It needs access to the local publishers that report events first.
A GNews response uses a small, conventional article object containing fields such as the title, description, content, URL, image, publication date, language, and source.
Webz.io returns those core article fields together with the data needed to analyze, rank, group, and filter the result:
| Data returned | Webz.io | GNews standard response |
| Full extracted article text | Yes | Paid plans |
| Article summary | Yes | Description |
| Author | Yes | Not in standard response |
| Crawl and update timestamps | Yes | Publication timestamp |
| Article sentiment | Yes | No |
| Entity extraction | Yes | No |
| Entity-level sentiment | Yes | No |
| Stock tickers | Yes | No |
| Categories and detailed topics | Yes | Limited headline categories |
| Publisher-domain rank | Yes | No |
| Source trust signals | Yes | No |
| Fake-news and satire indicators | Yes | No |
| Political-slant data | Yes | No |
| Syndication relationships | Yes | No |
| Social and engagement data where available | Yes | No |
Webz.io’s current product documentation describes full-text content, sentiment analysis, category and topic labels, named entities, source trust data, fake-news and satire indicators, political-slant tagging, and flexible source filtering.
In the broad-search benchmark sample, the median extracted Webz.io article text was 4,805 characters. The median visible GNews content value was 265 characters, and all 80 tested GNews content values were truncated because the supplied credential was on the free plan.
GNews paid plans include full content, so this is partly a plan comparison. The larger difference is that full text is only one part of the Webz.io response. Webz.io also supplies the enrichment and source context needed to use that content.
GNews supports keyword search, Boolean operators, title or content-field selection, language, publisher country, date ranges, and sorting by date or relevance. That is enough for many basic searches.
Webz.io adds controls that let developers express a complete monitoring policy in the query itself.
For example, an application can search for:
A simplified query could look like this:
import requests
response = requests.get(
“https://api.webz.io/api/news”,
params={
“token”: “YOUR_WEBZ_TOKEN”,
“q”: (
‘(title:(NVIDIA) OR text:(NVIDIA)) ‘
‘language:english ‘
‘country:US ‘
‘sentiment:negative ‘
‘ticker:NVDA ‘
‘domain_rank:<10000’
),
“size”: 100,
“sort”: “published”,
“order”: “desc”,
},
timeout=30,
)
response.raise_for_status()
data = response.json()
The benchmark directly validated several of these controls:
| Webz.io filter tested | Live result |
| site:yahoo.com | All ten returned records were on Yahoo domains |
| sentiment:negative | All ten returned records had negative sentiment |
| domain_rank:<1000 | All ten returned domains had ranks below 1,000 |
| ticker:NVDA | Returned matching ticker-based results |
| trust.top_news:top_news_us | Returned results from the selected top-news source group |
GNews currently says that filtering search results by source is not supported.
This difference becomes important when the application needs an allowlist, blocklist, approved publisher set, local-source policy, trust threshold, sentiment rule, or financial-entity filter.
With GNews, more of that logic must run after retrieval. With Webz.io, it can be applied before the articles enter the product.
A larger index will naturally contain more long-tail material. That is useful when recall matters, but most products do not want to present every possible mention directly to users.
Webz.io addresses this with server-side controls for:
This lets different products use the same source corpus in different ways.
A risk-monitoring product can favor recall and search full article text. A financial product can restrict results by company entity or ticker. A consumer feed can focus on top-ranked publishers. A regional monitor can preserve local reporting that a mainstream-only ranking would remove.
GNews offers a dedicated Top Headlines endpoint whose stories are selected using Google News rankings across nine categories. That is convenient for a conventional homepage or category feed.
Webz.io takes a different approach: it gives developers the coverage and the fields needed to build the ranking policy that fits their own product.
GNews supports familiar page-number pagination, but its documentation limits traversal to the first 1,000 articles returned by a query.
That may be enough for a news feed. It is more restrictive for:
Webz.io supports cursor-based pagination for collecting large result sets. It also provides historical open-web data going back to 2008. GNews’s paid historical archive begins in 2020, while its free plan is limited to 30 days.
In the benchmark’s 45-day search:
For a developer testing a one-week feed, that difference may not matter. For a product that needs backfills, long-term trend analysis, or event reconstruction, it does.
At the time of the freshness test, the newest returned article was:
| Provider | Age of newest article |
| Webz.io | 2 minutes 19 seconds |
| GNews | 12 hours 14 minutes |
This does not mean paid GNews data is always 12 hours behind. The tested GNews key was on the free plan, which explicitly includes a 12-hour delay. Paid plans advertise real-time availability.
It does show an important difference in how developers can evaluate the products.
Webz.io’s free account gives developers monthly API credit that can be used against its APIs. GNews’s free access is limited to development and testing, is not licensed for commercial use, and applies the 12-hour delay.
| Capability | Webz.io News API | GNews API |
| Stated news-source coverage | 350K+ sources | 80K+ sources |
| Stated language coverage | 170+ languages | 41 languages |
| Stated geographic coverage | 200+ countries | 71 countries |
| Title-only benchmark total | 21,328 | 951 |
| Broad-search benchmark total | 163,699 | 10,718 |
| Full extracted article text | Yes | Paid plans |
| Boolean search | Yes | Yes |
| Field-specific search | Extensive | Title, description and content |
| Phrase search | Yes | Yes |
| Wildcard and proximity search | Yes | Not documented |
| Publisher-domain filtering | Yes | Not currently supported |
| Domain exclusions | Yes | Not currently supported |
| Publisher-rank filtering | Yes | No |
| Sentiment filtering | Yes | No |
| Named-entity filtering | Yes | No |
| Stock-ticker filtering | Yes | No |
| Detailed topic taxonomy | Yes | No |
| Source trust signals | Yes | No |
| Fake-news and satire labels | Yes | No |
| Political-slant labels | Yes | No |
| Syndication data | Yes | No |
| Custom source requests | UI and API on eligible plans | Source-submission form |
| Dedicated top-headlines endpoint | Configurable through feed filters | Yes |
| Maximum query traversal | Large-scale cursor pagination | 1,000 articles |
| Historical archive | Back to 2008 | Back to 2020 on paid plans |
| Output | JSON or XML | JSON |
| Free commercial API usage | Monthly Webz.io credit | No |
| Free-plan real-time data | Yes with available credit | 12-hour delay |
The provider scale, product, and plan details in this table come from the providers’ current published pages.
GNews is a reasonable option when the requirement is narrow:
Its API has two main endpoints and a relatively small parameter set. Developers can make a basic request with little knowledge of news-data query design. GNews also offers official JavaScript, TypeScript, Node.js, and PHP libraries.
The tradeoff is that developers receive a smaller source universe and fewer tools for controlling, analyzing, and enriching the results.
Webz.io is the better fit when the application needs to:
The live benchmark found that Webz.io exposed far more searchable news while also returning the metadata needed to manage that coverage.
GNews is easy to understand: send a keyword query and receive a compact list of mainstream news articles.
Webz.io addresses the parts of news-data infrastructure that become difficult after the prototype:
In the live benchmark, Webz.io returned 22.4× more title matches and 15.3× more title-and-content matches. It won every paired coverage test and showed much wider reach in both multilingual probes.
For developers building a simple headline widget, GNews may be sufficient.
For developers building a news product, media-monitoring system, AI application, risk platform, or analytics pipeline where missing relevant coverage has a real cost, Webz.io is the better choice.
The benchmark contained 83 live test cases: 54 against Webz.io and 29 against GNews.
The main coverage matrix tested eight topics over the same 72-hour publication window. It included title-only and broad-content searches, English and non-English queries, publisher-country filters, Boolean syntax, freshness, historical access, pagination, page-size limits, and provider-specific filtering capabilities.
The benchmark deliberately ended the main coverage window more than 12 hours before testing so the GNews free-plan delay would not affect the coverage comparison.
Latency was not scored because the execution environment required an intermediary to collect the live response bodies. This preserved the API records, counts, pagination data, and plan notices, but it would not have produced a fair provider-native response-time comparison.
Reported result totals come from each provider. Indexing, deduplication, extraction, and query semantics can differ, so the title-only benchmark should be treated as the closest direct coverage comparison.
Download the benchmark package
Webz.io is better for products that require broad coverage, multilingual sources, full article text, historical data, advanced filtering, and enriched metadata. GNews is better suited to a simple mainstream news feed that does not need deep source control or downstream analytics.
Webz.io won every matched coverage query in the live benchmark. It returned 21,328 title matches compared with 951 from GNews, and 163,699 title-and-content matches compared with 10,718.
Webz.io also states coverage of more than 350,000 news sources, while GNews states coverage of more than 80,000.
Webz.io supports more than 170 languages across more than 200 countries. GNews markets 41 languages and 71 countries. In the live probes, Webz.io returned 59.1× more Hebrew results and 18.3× more Spanish results.
GNews paid plans include full content. Its free plan automatically truncates the content field. In the tested free-plan responses, all 80 sampled content values were truncated.
Webz.io returned full extracted article text in the tested responses and states that full-text content is included in its news data.
GNews currently states that source filtering is not supported. Its Search endpoint supports keywords, language, source country, dates, searchable fields, and result sorting.
Webz.io supports filtering and exclusion by publisher domain, full host, domain suffix, site section, source rank, trust type, and other source fields.
Webz.io is better suited to applications that need full article text, structured metadata, entities, categories, topics, sentiment, source information, and large-scale collection. Those fields can reduce the amount of extraction and enrichment work that must happen after retrieval.
For natural-language and semantic news retrieval, Webz.io also offers a separate News Search API that returns relevant article passages together with structured filters.
The pricing models are different.
GNews uses fixed monthly plans with daily request limits. Its free plan provides 100 requests per day, returns up to ten articles per request, has a 12-hour delay, and is restricted to development and non-commercial testing. Its commercial Essential plan currently starts at €49.99 per month.
Webz.io uses monthly free API credit, pay-as-you-go access, and custom high-volume packages. Every Webz.io account currently receives $5 in free API credit each month, with no credit card required.
Raw request prices should not be compared without considering the returned data. A Webz.io result can include full text, source metadata, sentiment, entities, taxonomies, trust data, and syndication fields that may otherwise require separate processing.
Customers on eligible Webz.io subscription plans can submit sources through a self-service UI or API. Webz.io automatically analyzes the source and returns a report on its availability and coverage status.
Search news from hundreds of thousands of sources, retrieve full article content, and control results by source, language, country, sentiment, category, entity, ticker, trust, and more.
Start free with $5 in monthly API credit. No credit card required.
Do you use Python? If so, this guide will help you automate supply chain risk reports using AI Chat GPT and our News API.
Use this guide to learn how to easily automate supply chain risk reports with Chat GPT and news data.
A quick guide for developers to automate mergers and acquisitions reports with Python and AI. Learn to fetch data, analyze content, and generate reports automatically.