01.DATA STREAMING// gRPC · shreds · order books
// gRPC · raw shreds · order-book websockets · one subscription

Chain state, pushed.
Millisecond delivery.

Supanode streams live state over the interface each network already speaks — Yellowstone-compatible gRPC and raw UDP shreds on Solana, native gRPC on Monad, order-book WebSocket on Hyperliquid. gRPC p99 of 3.8 ms, unmetered throughput. Built for snipers, MEV searchers, and HFT bots.

STREAMING · LIVE● operational
gRPC p99 · 60min
3.8 ms
▁▂▃▂▄▃▂▄▅▃▂▁▂▃▄▅
shred p99 · 60min
0.4 ms
▂▁▂▃▁▂▁▂▁▃▂▁▂▁▂▁
streams open
1,247
msg/s peak
48k
networks live
3
drop rate
0.02 %
02.WHERE IT RUNS// 4 live · 1 in build
03.SOLANA, IN DETAIL// 3 ways to stream
/ bundlesFLAGSHIP
Bundles
Pre-packaged gRPC + ShredStream in 5 throughput tiers. Best $/req for full-stack streaming consumers.
gRPCShredStream
⤷ one subscription
5 tiers · auto-scalesee in pricing →
/ shreds-udpRAW
ShredStream Raw
Raw UDP shred multicast direct from validator. The fastest possible data path — sub-millisecond from block.
validator
UDP multicast
your bot
standalone · for sniperssee in pricing →
/ shredstream-decodedSOON
ShredStream Decoded
Pre-parsed shreds as structured JSON. No reconstruction overhead — instant transactions and accounts.
validator
parsed
{ tx, accounts, … }
your bot
early accessjoin waitlist →
04.WHY OURS// 4 reasons
<5ms p99 stream latency
Measured at the edge, where your client connects.
Peering-dense European edge
Streaming nodes sit on one of the world's largest internet exchanges, with tight peering across Western & Central Europe (Amsterdam, London, Paris, Zurich), the Nordics, and into Eastern Europe & the Middle East. Other regions on request.
standard interfaces, no custom SDK
Every stream speaks what the chain already speaks — gRPC where the network has it, plain WebSocket where it does not. Same schemas, same subscribe API: point your existing client at our endpoint.
flat monthly, not per-request
Run 10× more bots without 10× the bill — the plan covers concurrent streams and message volume together, so a volatile week is not a surprise invoice.
05.CODE// solana example · subscribe in 5 lines
stream.tstypescript · yellowstone-grpc
1import Client from "@triton-one/yellowstone-grpc"
2
3const c = new Client("https://grpc.supanode.xyz", token)
4const stream = await c.subscribe()
5stream.on("data", msg => trade(msg.account))
~/supanodebash · grpcurl
$ grpcurl -H "x-token: $KEY" \
grpc.supanode.xyz:443 \
geyser.Geyser/SubscribeAccount
200 OK · stream open · ts 0.8ms
→ message · slot 287413298 · 0.4ms
→ message · slot 287413299 · 0.5ms
see full docs →examples for Python · Rust · Go · TS in /docs
06.PRICING// flexible billing

Try first.
Pay for what you need.

Up to 24h trial on any streaming product, no card. Flat monthly per product, with no per-message metering — on Solana you can bundle gRPC and ShredStream into one tier. Custom quote for >100k msg/s or dedicated streaming infrastructure.

// streaming pricing models
Trial→ /pricing
Bundles (Solana)→ /pricing/solana#bundles
ShredStream (Solana)→ /pricing/solana#shreds
Custom→ dedicated
07.GET STARTED// manual provisioning, fast

Streaming in 10 minutes.

Open Telegram, get your token, paste it into the client you already use. Done. Real engineers on call for the first hour of every onboarding.

→ opens Telegram · provisioning is manual
[ FAQ ]
What is the difference between gRPC and ShredStream?

Yellowstone gRPC delivers structured, filterable live state — accounts, transactions, slots, and blocks — over a Triton-compatible interface. ShredStream delivers the raw Solana shreds validators distribute over Turbine, with no filters, at the earliest possible point — ahead of any RPC indexing or gRPC re-emission. Use gRPC for filtered state; ShredStream for the lowest possible latency.

Is Supanode gRPC compatible with existing Yellowstone clients?

Yes — it is a drop-in replacement for Triton clients: same protobuf schemas, same subscribe API. Point your existing SDK at the Supanode endpoint and keep your code.

Is streaming throughput metered?

No. Throughput is unlimited and unmetered — there is no per-MB or per-credit billing. The tier sets concurrent connections and how many addresses you can watch in total; data volume on top of that is not billed.

How fast is delivery, and from where?

On Solana, about 3.8 ms p99 for gRPC and 0.4 ms p99 for ShredStream, measured at the edge where your client connects. See pricing for tiers and the per-IP ShredStream rate.

Which networks can I stream?

Solana over Yellowstone-compatible gRPC and raw UDP shreds, Monad over its own native gRPC feed, and Hyperliquid over the exchange WebSocket - all live today. A Lighter order-book stream is in build. Any other network is a dedicated node with the chain-native feed on it, usually deployed within 48 hours.