THE TEA · Content API
The Chinese tea encyclopedia, as an API.
The deepest Chinese-tea dataset, built for software. 871 teas and 157 topic articles, every fact typed and queryable — taste, brewing, terroir, chemistry, recipes, coordinates — plus semantic search and a built-in AI sommelier. 1028 articles, every one in all 72 languages. One request returns exactly what you need. No scraping, no half-translated wikis, no guessing.
Try it · live against api.thetea.app
Glossary · tea terms in your language
What you get
A normalized, typed dataset — not HTML blobs.
Granular fields
Fetch organoleptic.taste, brewing.water_temp, terroir.altitude directly — no scraping.
72 languages
1028 articles, every field, translated and validated for cross-language faithfulness — including Tibetan, Amharic, Bhojpuri.
Structured extras
Sensory profiles, gongfu and western brewing recipes, harvest calendars, comparison graph, coordinates.
Multilingual search
One endpoint matches names across Cyrillic, Han, pinyin and Latin.
Glossary
Canonical tea terms in 72 languages — for tooltips, localization, NLP.
Ready-to-use enrichment
Per tea: one-liner, summary, tasting note, food pairings, a 3-question FAQ, plus caffeine, difficulty and price tier — all 72 languages.
SEO out of the box
Localized seo_title, seo_description and keywords for every tea, in every language.
Provenance
Each value carries its source and review status. Honest, auditable data.
Intelligence
Not just data — understanding.
On-platform vector search and an LLM grounded in the dataset. Your users describe what they want in plain language; the API answers from real teas, with sources, in any language.
Semantic search
Find teas by meaning, not keywords — “smoky oolong with caramel”, “sweet low-caffeine green for the morning”. Filter by type or province; works in 72 languages.
AI sommelier
Ask a question, get a recommendation with the exact teas it used — grounded in the data, no invented facts. /api/v2/ask
Recommendations
“More teas like this” by sensory similarity — for cross-sell, discovery and tasting flights. /api/v2/tea/{slug}/similar
The API
A handful of endpoints. One base URL.
Every card now carries an enrichment block (summary, tasting note, pairings, FAQ, caffeine, difficulty) and a seo block — in all 72 languages.
Company directory
The companies behind the tea, worldwide.
The same API opens b2btea.com, a directory of 13,310 tea companies: factories, wholesalers, importers, exporters, tea houses, markets and cooperatives. Fifteen languages: English, Russian, Chinese, German, French, Spanish, Italian, Portuguese, Japanese, Korean, Turkish, Arabic, Polish, Dutch and Vietnamese.
Search
Filter by country, city, Chinese province, company type, tea, Chinese-tea focus and evidence level. A country or a type can be given as a key or in the request language: China, Chine, Китай. /api/v2/companies
Company card
What the b2btea.com page shows: description, address, phone, website, contact form, teas linked to encyclopedia articles, with a link back to the page. JSON or Markdown. /api/v2/company/{slug}
From a tea to its sellers
Start from an encyclopedia article and get the companies that deal in that tea. The directory tags companies by tea class and a few named teas; the answer says whether it matched the tea itself or its class. ?tea_article=pu-ercha
Counts
Companies by country, type, tea, region, Chinese-tea focus and evidence level, with labels in your language. /api/v2/companies/facets
No e-mail addresses, in any field. Entries come from open sources and do not claim that a company operates today; each carries its evidence level, from a website that answered a live check to a name a source mentioned. Without a key a search returns the first 10 results and no paging; company cards are always open.
AI-native
The same data, everywhere your users are.
MCP server
Tools for Claude Desktop, Cowork and any MCP client, including the b2btea.com company directory. Setup & download →
AI skill
Drop-in skill for Claude and GPT so an assistant answers from real data. Install →
llms.txt & OpenAPI
Chat bots
Telegram, Discord, Slack, LINE, WhatsApp, WeChat on the same backend. Set up →
Pricing
Free to explore. Paid to build.
Every plan reads the same dataset and every endpoint — semantic search and the AI sommelier included. Paid plans add languages, full field depth, bulk export and commercial rights. Less than a coffee a month.
Pro
Apps serving major markets
- 7 languages
- Full field depth
- 600 req / min
- Semantic + sommelier
Expert
Production & commercial
- All 72 languages
- Full field depth
- Bulk export
- Commercial use
Lifetime
Full access, locked in
- All 72 languages
- Full field depth
- Pay once — no subscription
- Commercial use
Lifetime is a one-time payment — $19.99 in June, $49.99 from July. Pay with card, Apple / Google Pay, Alipay, WeChat Pay and more. Preview pricing may change before the stable release. Authenticate with Authorization: Bearer <key>.