No deshredding pipeline to operate
Skip packet handling, reconstruction, and transaction extraction work. Shreder does the heavy lifting so your team can focus on application logic.
// Decoded Shreds
Subscribe to low-latency Solana transaction data extracted from shreds, with account filters and no raw deshredding infrastructure to maintain.
Decoded Shreds is ideal when you want early transaction data and simple filters. Choose Raw Shreds for packet-level control, Binary for even lower latency, or Fastlane/Geyser for post-execution state.
For public performance data, see Shreder’s shred-stream benchmarks. View benchmarks →
// Overview
Solana block data begins moving through the network as shreds before many downstream systems expose full processed updates. Raw shreds are powerful, but expensive to operate well: your team must receive packets, deduplicate, reconstruct entries, decode transactions, filter, and route the result.
Decoded Shreds moves that work into Shreder’s low-latency infrastructure. Instead of receiving raw UDP packets, your application subscribes to a decoded transaction stream with account filters.
Decoded Shreds is a pre-execution transaction stream for early intent detection. It does not include post-execution account state, logs, balances, or final outcomes. Use Geyser gRPC / Fastlane when you need that context.
// Why Decoded Shreds
Decoded Shreds removes the most operationally complex part of the raw path while keeping your application close to the data source.
Skip packet handling, reconstruction, and transaction extraction work. Shreder does the heavy lifting so your team can focus on application logic.
Subscribe only to relevant transactions using include, exclude, and required account filters.
Built from shred streams instead of waiting for Geyser updates.
For many teams, this is the right balance of speed, convenience, and operational cost.
// Use cases
Decoded Shreds is the right starting point when your application needs low-latency transaction visibility without raw shred reconstruction overhead.
Monitor specific markets, programs, or accounts and react to transaction intent without running a local deshredder.
Subscribe to filtered transaction streams for target accounts and route matches into opportunity-detection logic.
Track relevant program activity early, then pair with post-execution streams when users need final status.
Use Decoded Shreds as an early transaction source, then reconcile with Fastlane, Geyser, or RPC for full execution metadata.
For even lower latency consider using Binary.
// Benchmarks
We benchmarked Decoded Shreds against other shred-stream providers in the same region and setup. Decoded Shreds delivered the first actionable signal in the vast majority of cases—without the operational overhead of running your own decoder.
Run 001
Measured in Frankfurt, Teraswitch FRA2.
| Provider | p50 | p95 | p99 |
|---|---|---|---|
| Shreder.xyz Frankfurt | 0.00 ms | 0.00 ms | 0.00 ms |
| Jito Frankfurt | 4.61 ms | 38.51 ms | 51.12 ms |
// How it works
01
Before the downstream APIs expose a transaction, Solana block data is already moving through the network as shreds.
02
Shreder infrastructure receives shreds in low-latency regions, reconstructs transaction data, and prepares updates for subscriptions.
03
Your client sends include/exclude/required account rules and Shreder matches updates server-side.
04
Matching transactions stream to your service. You can also confirm outcomes with Fastlane/Geyser when needed.
// Filters and format
Decoded Shreds supports account-based transaction filters so your hot path receives less noise before business logic.
account_includeMatch transactions involving any account in the include list.
account_excludeIgnore transactions involving any account in the exclude list.
account_requiredMatch transactions that include all required accounts.
{
"pumpfun": {
"account_include": [],
"account_exclude": [],
"account_required": [
"6EF8rrecthR5Dkzon8Nwu78hRvfCKubJ14M5uBEwF6P"
]
}
}
// Performance
For production systems, the useful latency metric is when a transaction becomes actionable in your app.
A raw feed may reach your server earlier, but if your local decoder/filter path is slower than a managed decoded stream, raw can lose the end-to-end race.
Use public benchmarks as a starting point, then run controlled tests from your production region and hardware profile. View shred stream benchmarks →
// Benchmarking
Use an open-source benchmark to measure Decoded Shreds latency in your current setup and compare it side by side with other providers.
GeyserBench is an open-source tool for realistic benchmarking from your own environment. It lets you run the same monitored accounts and filters against multiple endpoints and compare results under your actual network path and hardware profile.
The output helps you validate first-detection share and latency deltas (p50/p95/p99) for your current production setup, so you can decide whether Decoded Shreds improves your real time-to-actionable signal.
// Compare
Decoded Shreds is the practical starting point for many teams because it combines shred-level delivery with transaction-level ergonomics.
| Product | Format | Best for | Customer-side work | Latency profile |
|---|---|---|---|---|
| Raw Shreds | Raw UDP shred packets | Teams with custom deshredding and strict latency budgets | Highest: receive, dedupe, reconstruct, decode, filter | Earliest feed; end-to-end speed depends on your decoder |
| Decoded Shreds | Structured transaction stream with filters | Production teams that want shred-level delivery without running a decoder | Low: subscribe and process transaction updates | Low latency with less operational overhead |
| Binary | Transaction data from shreds | Trading systems that want efficient transaction-level delivery | Low: no local decoding required | Lowest-latency transaction-level path for most workloads |
| Geyser gRPC / Fastlane | Yellowstone-compatible gRPC | Existing Geyser users, indexers, wallets, DeFi apps | Low: use existing gRPC tooling | Same interface; Fastlane is optimized for speed |
Decoded Shreds is pre-execution transaction data. For account state, logs, balances, and execution results, use Geyser gRPC or Fastlane.
Not sure which stream to use? Talk to Shreder →
// Requirements
Decoded Shreds is simple to integrate, but production systems should still handle stream lifecycle, reconnection, and backpressure.
Decoded Shreds delivers transaction-level data extracted from Solana shreds while Shreder handles low-level reconstruction and decoding.
Raw Shreds delivers unprocessed UDP packets and requires your own deshredding stack. Decoded Shreds moves that work into Shreder and streams transaction updates directly.
Decoded Shreds returns structured transaction updates. Binary returns serialized transaction payloads that you deserialize locally. Both provide decoded transaction data, however Binary gives further latency improvements. Consider using Binary for workloads with even stricter latency requirements.
No. Decoded Shreds is part of Shred Streams and focuses on early pre-execution transaction data. Fastlane/Geyser is for Yellowstone-compatible post-execution workflows.
No. For post-execution context, use Fastlane or Geyser gRPC.
No. Decoded Shreds is built for teams that want early transaction data without operating a raw decoder pipeline.
account_include, account_exclude, and account_required.
Yes, you can set up multiple filters and multiple connections.
Frankfurt, Amsterdam, New York, London, and Tokyo.
Use GeyserBench to compare Decoded Shreds with other providers in your region and setup. See the Benchmarks page.
Choose Raw Shreds when you need packet-level control and are ready to run a high-performance deshredding pipeline.
Start from the Shreder console or contact the team. We will provide endpoint and region guidance.
Tell us your target region, accounts or programs to monitor, current provider, and latency goals. We will help you configure Decoded Shreds and benchmark it against your setup.