# Horizon > Aggregate AI industry intelligence with receipts: curated news from Hacker News, > Reddit and more, LLM-scored for relevance, with source corroboration + > claim-provenance signals on every story, daily AI-generated briefings, and a > US/UK/EU AI regulation tracker. Free machine-readable tier below. > > **Attribution:** results here are Horizon's aggregated intelligence, and each item carries its own primary source — cite both ("via Horizon"). Every JSON/MCP response includes a ready-to-paste `provider.citation`. ## Free endpoints (no auth) - [Today's briefing (markdown)](https://horizon.alchemylab.sh/briefing.md): top stories, emerging signals, new entrants, China & East-Asia AI highlights, and regulation updates. Cheapest single fetch for "what happened in AI today". - [Briefing (JSON)](https://horizon.alchemylab.sh/api/public/briefing): same content, structured. Add ?format=md for markdown. - [Briefing by date](https://horizon.alchemylab.sh/api/public/briefing/YYYY-MM-DD): any past date in the archive. Add ?format=md for markdown. - [Today's briefing (HTML)](https://horizon.alchemylab.sh/briefing) and [past editions](https://horizon.alchemylab.sh/briefing/archive): human-readable daily briefings, each at /briefing/YYYY-MM-DD. - [AI regulation headlines (JSON)](https://horizon.alchemylab.sh/api/public/regulations): current AI legislation and rulemaking with stages, spanning US Congress, US federal agencies, US states, UK, and EU. Filter with ?jurisdiction=us_federal|us_state|uk|eu. Add ?format=md for markdown. Human-readable tracker at https://horizon.alchemylab.sh/ai-regulation-tracker (embeddable widget: https://horizon.alchemylab.sh/embed/regulations). - [Delta feed (JSON)](https://horizon.alchemylab.sh/api/public/feed): `?since=&limit=®ion=china` — poll for newly-added items without re-fetching. Returns `{items, nextCursor, hasMore}`; pass nextCursor back as `since` each poll. Add `region=china` for only China/East-Asia AI items. Metadata only (titles, links, topics, scores) — full bodies and higher limits are on /api/v1/feed. - [Related items (JSON)](https://horizon.alchemylab.sh/api/public/related): `?id=&limit=` — the most semantically similar items to a given one (ids come from the feed/search). Metadata only; full bodies on /api/v1/related. - [RSS](https://horizon.alchemylab.sh/feed.xml): daily briefings feed. - [How verification works](https://horizon.alchemylab.sh/methodology): credibility tiers, corroboration levels, and claim-provenance tags (vendor-claimed vs independently evaluated) explained. - Topic hubs at https://horizon.alchemylab.sh/topics/ (models, tools, use-cases, regulation, research, funding, opinion, tutorial) and entity hubs at https://horizon.alchemylab.sh/entities/ — recent relevance-scored coverage per topic/company/model/person. - Regional AI hubs: [China](https://horizon.alchemylab.sh/china) (Qwen, DeepSeek, Zhipu, Moonshot, ModelScope), [Korea](https://horizon.alchemylab.sh/korea) (HyperCLOVA X, EXAONE, Naver, Samsung), [Japan](https://horizon.alchemylab.sh/japan) (Sakana AI, Preferred Networks, NTT tsuzumi) and [Europe/EU](https://horizon.alchemylab.sh/eu) (Mistral, Aleph Alpha, Black Forest Labs, the EU AI Act) — relevance-scored coverage per region, **including local-language press (Chinese, Korean, Japanese, German) translated into English**; also a `region=china|korea|japan|eu` filter on the feed/search and the get_region_signal MCP tool. ## MCP server Streamable HTTP endpoint at https://horizon.alchemylab.sh/api/mcp. No auth for the read tools. Per-host setup (Claude web/desktop custom connector, Claude Code, Cursor, VS Code, Windsurf, Cline, ChatGPT developer mode, OpenAI Responses API) with verified config for each: https://horizon.alchemylab.sh/developers Free tools (generated from the server's own definitions): - get_daily_briefing — Today's AI briefing, or any past day's: top stories with their sources and corroboration labels, emerging signals, new entrants, and regulation updates. - get_regulation_updates — Current AI legislation headlines across the US Congress and federal agencies, US state legislatures, the UK Parliament, and the EU, with each item's legislat... - get_china_signal — Shortcut for get_region_signal with region "china": the current China & East-Asia AI signal (Qwen, DeepSeek, Zhipu, Moonshot, ModelScope, and more) that West... - get_region_signal — The current AI signal for a region (china, korea, japan, or eu) — recent, relevance-scored items on that region's models, labs, and analysis, ranked by momen... - get_entity_provenance — For one company, model, or person: how much of the claim activity Horizon holds provenance data for is self-reported by that entity versus independently eval... - get_blind_spots — Coverage blind spots: companies, models, and people that regional press (Chinese/East-Asian, European, Latin-American) is covering while Western/global feeds... - get_topic_signal — The current AI signal for a topic — recent, relevance-scored items tagged with that topic, ranked by momentum. - search_news — Quick title search over the last 14 days of AI news: returns the top 3 matches by momentum, with links and item ids (pass one to get_related). - get_new_since — Incremental poll: raw item-level AI news added since a cursor, oldest→newest, with a nextCursor for your next call — use this for "what's new since I last ch... - get_related — Semantic 'more like this': given an item id (from get_new_since, search_news, or the feed), returns the most similar recent items by embedding similarity. - list_sources — Describe what Horizon covers: ingestion sources with their source= filter values, the regional lenses (region= / dedicated tools), and how to filter. Two more need a Horizon API key sent as `Authorization: Bearer hzn_live_...`: - ask_horizon — Ask a natural-language question about recent AI developments and get a synthesized, cited answer drawn only from Horizon's corpus — one call instead of searc... - get_trends — The week's rising entities in AI: the companies, models, and people whose mentions accelerated most over the last 7 days versus the prior week — momentum, no... (Pro plan) ## Paid agent API (API key) Subscribers can mint API keys at https://horizon.alchemylab.sh/api-keys for the /api/v1 tier — full-depth access, authenticated with `Authorization: Bearer hzn_live_...` (120 req/min; monthly quota by plan — Standard 10k, Pro 100k — a 429 with `resetsAt` when exceeded): - **Machine-readable spec**: [OpenAPI 3.1](https://horizon.alchemylab.sh/openapi.json) — also at `/.well-known/openapi.json`, with an ai-plugin manifest at `/.well-known/ai-plugin.json`. Import it into GPT Actions, LangChain, or any OpenAPI tool-caller to auto-generate a typed client for the endpoints below. - [Ask Horizon](https://horizon.alchemylab.sh/api/v1/ask): POST JSON `{"question": "what did OpenAI ship this week?", "days": 14, "limit": 18}` — returns a synthesized, cited answer grounded only in Horizon's recent corpus: `{answer, citations:[{n,title,url,source,publishedAt}], retrieved}`. Answer text cites sources inline as [n]. One call instead of search-then-read. Runs an LLM per call, so it has a tighter monthly allowance than the read endpoints (see /api/v1/usage). - [Full search](https://horizon.alchemylab.sh/api/v1/search): `?q=&source=&topic=®ion=china&from=&to=&limit=&offset=&collapse=` — searches all content (not the 3-item free teaser). Near-duplicate stories are collapsed to one representative per cluster by default (each carries `clusterSize`); pass `collapse=false` for every matching row. - [Delta feed](https://horizon.alchemylab.sh/api/v1/feed): `?since=&limit=®ion=china` — same incremental poll as the free feed but with full bodies and up to 200 items/page. - [Related items](https://horizon.alchemylab.sh/api/v1/related): `?id=&limit=` — full-body "more like this" over the semantic embeddings. - [Bulk briefing archive](https://horizon.alchemylab.sh/api/v1/briefings): `?from=&to=&limit=` — the whole history in one range query, full bodies included. (Individual days are free at /api/public/briefing/YYYY-MM-DD and /briefing/YYYY-MM-DD; this is the bulk/export convenience.) - [Webhooks](https://horizon.alchemylab.sh/api/v1/webhooks): register your own push endpoints. POST JSON `{"url":"https://...","topics":[],"entities":[],"keywords":[]}` (signing secret returned once); GET lists them; DELETE /api/v1/webhooks/ removes; POST /api/v1/webhooks//test fires a signed test event. Same as the dashboard, key-gated. - [Usage](https://horizon.alchemylab.sh/api/v1/usage): your key's current rate-limit + monthly-quota state `{rateLimit:{limit,remaining,resetSeconds}, quota:{limit,used,remaining,resetsAt}}`. Every /api/v1 response also carries `X-RateLimit-*` and `X-Quota-*` headers, so you can self-throttle without a separate call. ## Webhooks (push) Subscribers configure webhooks at https://horizon.alchemylab.sh/webhooks to have newly-matched items (by topic/entity/keyword, or all) POSTed to their endpoint in near-real time. Each delivery is signed: `X-Horizon-Signature: t=,v1=` where the HMAC-SHA256 key is your signing secret and the message is `.`. Verify it and reject stale timestamps. Agents can self-register and manage webhooks programmatically via the key-gated /api/v1/webhooks endpoints above. ## Paid tier ($5/mo) Live scored feed, full-text search, complete regulation timelines and history: https://horizon.alchemylab.sh/subscribe. Free daily email digest: https://horizon.alchemylab.sh/newsletter ## Policies - [Privacy](https://horizon.alchemylab.sh/privacy) - [Terms](https://horizon.alchemylab.sh/terms)