FXMacroData vs Marketaux: Financial News API vs Structured Macro Release Data

This comparison is for FX traders, quant developers, and AI-agent builders who are deciding whether a financial news API with entity tagging and sentiment can stand in for structured macro release data. It is a common shortcut: subscribe to a news feed, filter for "CPI" or "payrolls", and treat the first matching headline as the release. Marketaux makes that shortcut look easy, because it tags articles with entities, countries, and sentiment scores out of the box.

The core finding is that the two products answer different questions. Marketaux tells you what the financial press is saying about an entity and how positive or negative that coverage reads. FXMacroData tells you what the official release actually printed, when it was published to the second, and what the prior value was, as a structured row a program can trust. Headlines are a useful signal about reaction; they are not the data. If your workflow trades or models the release itself, FXMacroData is the better fit; if it trades or monitors the narrative around companies and markets, Marketaux is the more natural tool.

Decision snapshot

Choose Marketaux when

You need a broad financial news feed with entity tagging across stocks, ETFs, indices, currencies and crypto, per-entity sentiment scores, 30+ languages, and a free tier for prototyping.

Choose FXMacroData when

You need the official macro print itself: value, prior, second-level announcement timestamp, release calendar, and FX context across 22 currencies, via REST or MCP.

Fast verdict

Marketaux wins for news monitoring and sentiment features. FXMacroData wins for release-driven FX trading, event studies, and agents that need numbers rather than headlines.

Takeaway: Pick by the object your system acts on: an article with a sentiment score, or a timestamped official release row.

Pricing And Plans

Marketaux publishes clear self-serve pricing. Its pricing page lists a Free plan at $0* with 100 requests per day and 3 articles per request, Basic at $29/month* (2,500 requests/day, 20 articles per request), Standard at $49/month* (10,000 requests/day, 50 articles per request), Pro at $99/month* (25,000 requests/day, 100 articles per request), and Pro 50K at $199/month* (50,000 requests/day). Annual billing lowers each paid tier (for example Standard is $41/month* billed annually), and custom plans with custom SLAs are available on request.

FXMacroData has an Always Free tier with no key required (USD data only, 100 requests/day, USD announcements delayed 15 minutes), Individual at $50/month or $500/year with a 14-day trial and 1,000,000 requests/month, Business at $250/month or $2,500/year for companies and teams of up to 10 people, and Enterprise by custom agreement. A self-serve Commercial Redistribution add-on costs $10/month per block of up to 100 measured users or 1,000 public-feed followers.

On sticker price the two are close: Marketaux Standard and FXMacroData Individual sit within a dollar of each other. The difference is what the money buys. Marketaux Standard caps at 10,000 requests per day, roughly 300,000 per month; FXMacroData Individual includes 1,000,000 requests per month with bursts of 300/minute. And the payload is different: a Marketaux request returns article snippets with entity and sentiment metadata, while an FXMacroData request returns official release values with timestamps.

Monthly list price, USD

FXMacroData in mint; Marketaux plans in cyan. Hover a bar for the exact price.

Takeaway: Prices are in the same band, so the decision rests on the data type: news snippets with sentiment, or official macro releases with second-level timestamps and a 1,000,000 requests/month allowance.

FXMacroData Individual

$50/month

1,000,000 requests/month, 22 currencies, 700+ official indicator series, release calendar, COT, FX rates, MCP. 14-day trial.

Marketaux Standard

$49/month*

10,000 requests/day, 50 articles per request, entity stats, sentiment, 5,000+ sources, 30+ languages.

Free tiers

$0

Marketaux: 100 requests/day, 3 articles per request*. FXMacroData: 100 requests/day, USD only, no key, 15-minute delay on USD announcements.

Takeaway: Both are easy to evaluate without a sales call; neither price gap is large enough to decide the question on its own.

Side-By-Side Comparison

* Competitor pricing retrieved from their public pricing page on October 2026.

Why Headlines Are Not Release Data

The appeal of a news API for macro trading is that it is general: one feed covers earnings, mergers, central banks, and data releases. Marketaux's entity model is genuinely capable. Each article carries entity objects with a symbol, type, country, a match_score for relevance, and a sentiment_score between -1 and +1 that Marketaux describes as the average sentiment of the highlighted text mentioning that entity.

The problem for a release workflow is that none of those fields is the release. When US CPI prints, a news article about it is a secondary document: it was written after the release, its timestamp is the article's publication time, the number lives inside prose or a snippet, and its sentiment score measures the tone of the writing, not the surprise versus expectations. A "negative" CPI article could mean inflation was hotter than expected (bad for bonds, often good for the dollar) or that a writer framed a soft print pessimistically. Sentiment tone and macro surprise are different variables.

Same event, two different objects

News article (Marketaux shape)

• title, snippet: prose written by a publisher.
• published_at: when the story went out, not when the data did.
• entities[].sentiment_score: tone of the text, -1 to +1.
• match_score: how relevant the entity is to the article.
• The printed number must be parsed out of text, if it is there at all.

Release row (FXMacroData)

• val: the official printed value as a number.
• prior and pct_change: ready for surprise logic.
• announcement_datetime: official publication time, to the second.
• Source metadata per series for auditability.
• Same schema across 22 currencies.

Takeaway: A sentiment score tells you how a story reads. A release row tells you what the economy printed and exactly when the market could know it.

Detection is the second problem. To use news for release detection you need to search for terms, filter by country and entity, de-duplicate syndicated stories (Marketaux offers a group_similar option for this), and then extract a number reliably from snippets that vary by publisher and language. Every one of those steps can fail silently, and the fastest article on a release is often a short wire headline that may not be among the sources you are filtering. With structured release data the detection step disappears: the row exists or it does not.

Release Timing And Latency

Marketaux describes its news as instant and its entities as tracked every minute, which is a fair description of a news pipeline. But a news pipeline can only be as fast as the publishers writing the stories, and then its own ingestion on top. For scheduled macro data, the clock that matters starts at the official publication by the statistics agency or central bank.

FXMacroData anchors every release row to a second-level announcement_datetime, and releases are typically available within seconds of official publication. The target is one second from publication, and the measured record is public on the release-speed page, broken down by source, with every release that took longer listed. That makes the latency claim checkable rather than a marketing line.

Where each product enters the release timeline

1. Scheduled

The release calendar shows the time and country in advance.

2. Official publication

The agency publishes; FXMacroData's row typically appears within seconds.

3. Wire and press

Publishers write headlines and articles about the number.

4. News API

Marketaux ingests, tags entities, and scores sentiment on those articles.

Takeaway: News sits downstream of the release by construction, so it is better for measuring reaction than for detecting the print.

Coverage: Entities Versus Indicators

Marketaux's coverage is measured in entities and sources: 200,000+ entities, 5,000+ sources, 80+ markets, 30+ languages, spanning equities, indices, ETFs, mutual funds, futures, commodities, currencies, and crypto. That breadth is the main reason to choose it. For a multilingual news dashboard or a screener that flags companies with a sudden swing in coverage tone, it is a broad, inexpensive base.

FXMacroData's coverage is measured in currencies and indicator series: 22 currencies and 700+ series covering policy rates, CPI and core CPI, GDP, non-farm payrolls and labour data, PMIs, retail sales, trade, current account, and wages. Around those series sit the datasets an FX desk uses daily: FX spot rates, CFTC COT positioning, government bond yields, commodities, central-bank press releases from the Federal Reserve, ECB, Bank of Japan and Bank of England, FX sessions, and pair dashboards such as EUR/USD and USD/JPY.

API Design And Query Model

Marketaux's query model is search-shaped. You call /v1/news/all with filters such as symbols, countries, entity_types, industries, sentiment_gte/sentiment_lte, min_match_score, must_have_entities, language, and date bounds, and you get a page of articles. Entity endpoints add intraday and aggregated sentiment statistics and trending entities. Authentication uses an api_token query parameter.

FXMacroData's query model is series-shaped: one endpoint per currency and indicator returns the release history. Authentication goes in the header so keys stay out of logs and URLs:

curl -H "X-API-Key: YOUR_API_KEY" "https://api.fxmacrodata.com/v1/announcements/usd/non_farm_payrolls"

Example response (abridged, illustrative values):

{
  "data": [
    {
      "date": "2026-08-01",
      "val": 142.0,
      "prior": 114.0,
      "pct_change": 24.56,
      "announcement_datetime": 1788525000
    }
  ]
}

Fields that matter

• announcement_datetime is epoch seconds of the official release.
• val and prior are numbers, not text to parse.
• pct_change is ready for surprise and filter logic.

The equivalent with a news API

Search for payroll stories, filter by country, group syndicated duplicates, then extract a number from snippet text and hope the first story was the fastest.

Takeaway: One call returns a machine-ready history; the news route needs a parsing pipeline before you have a single number.

Sentiment, AI Agents, And MCP

Sentiment is where Marketaux has a feature FXMacroData does not try to offer. Per-entity sentiment and aggregated entity statistics let you chart how coverage tone for a currency, index, or company changes through the day. That can be a useful secondary feature in a model, for example as a crowding or narrative indicator alongside price.

For AI agents the question is what the agent can cite. FXMacroData runs an official MCP server, so Claude, ChatGPT, Cursor and other MCP clients can query releases, the calendar, FX rates, and COT directly and quote the official number with its timestamp. Marketaux does not list an official MCP server; a third-party community wrapper exists that exposes its news search. An agent asked "what did the Fed policy rate decision say?" is on firmer ground citing a release row and the central bank's own statement than summarising a set of headlines. See AI trading with FXMacroData for agent patterns.

Where Marketaux Is Genuinely Stronger

• Breadth of news: 5,000+ sources and 200,000+ tracked entities across equities, crypto, commodities and currencies.
• Sentiment out of the box: per-entity sentiment scores, highlights, and aggregated entity statistics with sentiment filters in the query.
• Languages and markets: 30+ languages and 80+ markets suit international news products.
• Low entry price: Basic at $29/month* is below FXMacroData Individual, and annual billing reduces every tier.
• Narrative monitoring: for stories that are not scheduled data, such as M&A, guidance, or geopolitical events, a news API is the right tool and a release API is not.

Where FXMacroData Is Stronger

FXMacroData pros

• Official values with second-level announcement timestamps.
• Public, per-source release-speed record.
• 22 currencies in one schema, plus calendar, COT, yields, FX rates.
• 1,000,000 requests/month, header auth, official MCP.

FXMacroData tradeoffs

• No news feed or sentiment scoring.
• Focused on macro and FX, not single stocks or crypto.
• Free tier limited to USD with a 15-minute delay.

Marketaux pros

• Broad multilingual news with entity tagging.
• Sentiment filters and entity statistics.
• Published pricing from $29/month* and a free tier.

Marketaux tradeoffs

• No economic calendar or structured macro values.
• Snippets and links only, not full article text.
• Daily request caps; token passed as a query parameter.
• No official MCP server listed.

Takeaway: Marketaux is a capable news and sentiment layer. FXMacroData is the stronger choice when a system needs to act on the official number and its timestamp.

Decision Matrix

Recommendation

Choose Marketaux if your product is about news: monitoring what is being written about companies, sectors, currencies, or crypto, scoring its tone, and surfacing it in many languages. Its pricing is published, its free tier is enough to prototype, and its entity and sentiment filters are well suited to that job.

Choose FXMacroData if your workflow is about the release: trading or modelling unemployment, payrolls, inflation, or ECB policy rate decisions, joining them to pair moves, and giving an agent numbers it can cite. Compare plans, start with the API reference, check the measured release speed, or start the 14-day trial. If you are weighing other news-first tools, the FXMacroData vs Finlight comparison covers a news API with a different design.

Bottom line: Marketaux turns financial news into tagged, sentiment-scored articles. FXMacroData turns official macro releases into timestamped rows. At similar prices, pick the one whose output your system actually acts on, and for release-driven FX workflows that is the release row, not the headline.

Takeaway: Use news for reaction and narrative; use structured release data for the print itself.

Sources Checked

Marketaux pricing

Plan names, monthly and annual prices, daily requests, articles per request.

View source

Marketaux documentation

Endpoints, filters, entity fields, sentiment definition, authentication.

View source

Marketaux FAQ

Free plan, annual billing, cancellation, snippet-only article content.

View source

FXMacroData

Plans, API reference, measured release speed, MCP server.

Pricing · Release speed

Takeaway: Every Marketaux figure in this article comes from its public pricing, documentation, and FAQ pages, checked in October 2026.

Related Reading

• FXMacroData vs Finlight
• FXMacroData vs Finnhub
• FXMacroData vs Alpha Vantage
• Best economic indicator feeds for algo trading