Documentation
API & Setup
TON Radar runs in real-API mode — all endpoints proxy live data from STON.fi, DeDust, and TON Center. No mocks, no fixtures.
Quick start
cp .env.example .env.local # set TONCENTER_API_KEY to avoid rate limits npm install npm run dev # open http://localhost:3000
Environment variables
TONCENTER_API_KEY= TONCENTER_API_BASE=https://toncenter.com/api/v3 STON_API_BASE=https://api.ston.fi DEDUST_API_BASE=https://api.dedust.io TON_ACTION_LOOKBACK_HOURS=24 NEXT_PUBLIC_SITE_NAME=TON Radar
Internal endpoints
GET
/api/healthService health probe.GET
/api/pairs/new?limit=80Recently active DEX pairs across STON.fi and DeDust.GET
/api/tokens/new?limit=60Latest jetton deploys and mints from TON Center v3.GET
/api/pairs/:address?dex=DeDustFetch a single normalized pair.GET
/api/pairs/:address/trades?dex=DeDust&limit=80Live trade feed for a pool.GET
/api/pairs/:address/candles?dex=DeDust&interval=1mOHLCV candles derived from real swaps.GET
/api/search?q=TONSearch across tokens, pairs, and contract addresses.GET
/api/ton/actions?action_type=jetton_swap&limit=100Raw decoded TON Center actions.Upstream data sources
Going production
This starter intentionally ships without a database — proxy mode is enough to validate the data shape. To reach Dexscreener-class latency and history, add a persistent indexer that polls these endpoints every 1–5 seconds and writes tokens, pairs, swaps, liquidity_events, and candles into PostgreSQL/TimescaleDB or ClickHouse, then push updates to the UI over WebSocket or SSE.