Developer API

Everything Modelwire publishes — original headlines, 80–120 word summaries, editorial analysis, buzz scores, and links to the original reporting — as clean JSON. Free while in beta.

Quickstart

Pass your key in the x-api-key header. Base URL: https://api.themodelwire.com/v1/public

curl -H "x-api-key: mw_live_..." \
  "https://api.themodelwire.com/v1/public/articles?category=models&limit=5"

// JavaScript
const res = await fetch(
  "https://api.themodelwire.com/v1/public/articles?limit=5",
  { headers: { "x-api-key": process.env.MODELWIRE_API_KEY } },
);
const { articles, nextCursor } = await res.json();

# Python
import os, requests

r = requests.get(
    "https://api.themodelwire.com/v1/public/articles",
    headers={"x-api-key": os.environ["MODELWIRE_API_KEY"]},
    params={"category": "research", "limit": 10},
)
articles = r.json()["articles"]

Endpoints

GET/articlesPublished articles, newest first. Filters: category, source, q (text search), before (cursor), limit.

GET/articles/topCurrent top stories by buzz score (limit ≤ 8).

GET/articles/{id}One article plus up to 3 related articles.

GET/categoriesCategory catalog with live article counts.

GET/sourcesThe source catalog we aggregate from.

GET/briefingToday's Daily Landscape synthesis.

Pagination: pass the nextCursor value from a list response as before on the next request. before accepts unix milliseconds or an ISO 8601 timestamp. The full contract is in the OpenAPI 3.1 spec.

The article object

{
  "id": "01KWMQK8T3666MYKAQYYJXD32P",
  "headline": "Microsoft consolidates Copilot into unified super app...",
  "sourceHeadline": "Microsoft follows Anthropic and OpenAI into...",
  "sourceUrl": "https://techcrunch.com/...",
  "modelwireUrl": "https://themodelwire.com/article/microsoft-consolidates-...",
  "summary": "80-120 words of original synthesis...",
  "whyItMatters": "One-line strategic angle (top stories only)" ,
  "analysis": {
    "voice": "analyst",
    "whatsNew": "...", "howThisFits": "...", "whatToWatch": "...",
    "generatedAt": "2026-07-03T19:39:26.204Z"
  },
  "source": { "slug": "techcrunch", "name": "TechCrunch" },
  "categories": ["products", "business"],
  "entities": ["Microsoft", "Copilot"],
  "buzzScore": 78,
  "publishedAt": "2026-07-03T19:24:40.000Z",
  "updatedAt": "2026-07-03T19:39:26.204Z",
  "imageUrl": "https://..." 
}

headline and summaryare Modelwire’s original writing. sourceHeadline and sourceUrl belong to the original publisher. whyItMatters, analysis, and imageUrl are null when absent.

Rate limits & pricing

Free (beta): 1,000 requests per day per key, max page size 50. Quotas reset at midnight UTC. Every response carries x-ratelimit-limit, x-ratelimit-remaining, and x-ratelimit-reset headers; exceeding the quota returns 429.
Paid tiers are planned. When they launch, we will email every key holder at least 30 days in advance, and we intend to keep a free tier for hobby and evaluation use. Nothing changes without notice.

Attribution requirements

If you display API content, you must:

Link to the original reporting (sourceUrl) wherever you show a summary or headline — the journalism belongs to the people who did it.
Credit Modelwire with a link (to modelwireUrl or themodelwire.com) where our summaries or analysis appear.
Not present our summaries as full articles or use them to substantially reproduce a publisher’s content.

Terms

The API is provided as-is during beta, with no uptime guarantee (it has been reliable, but we are a small operation). Keys are personal; don’t ship them in client-side code you don’t control. Abuse, resale of raw feed data, or scraping around the rate limit gets a key revoked. Our standard terms of use and privacy policy apply. Questions, higher limits, or feedback: hello@themodelwire.com.