Solana Nodes and Low-Latency Feeds
OnchainDivers provides managed Solana data infrastructure for applications that need anything from a standard filtered gRPC stream to the earliest available transaction-execution signals. All prices below are monthly in USD.
Node access
| Service | Topology | Price | Best for |
|---|---|---|---|
| Basic gRPC | Shared | $200/month | Indexers, dashboards, account monitoring, and general real-time transaction feeds |
| Very-fast gRPC | Shared | $700/month | Latency-sensitive production consumers that still want shared infrastructure |
| Ultra-low-latency gRPC | Shared | $1,200/month | Trading systems that need approximately 1 ms lower delivery latency than the $700 tier |
| Dedicated custom gRPC | Dedicated | $3,300/month | The $1,200 low-latency stack on an isolated node with customer-specific configuration |
Basic shared gRPC — $200/month
The shared gRPC service provides filtered transactions, account updates, slots, blocks, and block metadata without requiring customers to operate validator-grade hardware. It is the practical default for indexers and applications that need a reliable push stream instead of polling JSON-RPC.
Very-fast shared gRPC — $700/month
This is the faster shared gRPC tier for latency-sensitive production consumers. It retains shared economics while using the lower-latency delivery path intended for trading and time-sensitive ingestion.
Ultra-low-latency shared gRPC — $1,200/month
This shared tier reduces delivery latency by approximately 1 ms compared with the $700 service. It is intended for systems where a millisecond is materially important but isolated node capacity is unnecessary.
Dedicated custom gRPC — $3,300/month
The dedicated tier runs the $1,200 low-latency stack on an isolated Solana node with customer-specific configuration. Capacity is not shared with unrelated customers, making this the right choice for sustained high-volume ingestion and production trading systems that need predictable resources and custom tuning.
Earliest-data feeds
| Feed | Price | What it exposes |
|---|---|---|
| Shreds | $500/month | Raw Turbine data and coding packets for reconstructing entries, transactions, and blocks before conventional RPC delivery |
| Preconfirmations | $1,600/month | Transaction execution-status signals emitted before shred propagation and standard RPC commitment levels |
Shreds — $500/month
Solana leaders split block data into network-sized packets called shreds. Data shreds carry block contents, while coding shreds add erasure recovery so a receiver can reconstruct a block even when some packets are missing. Consuming the shred stream removes the wait for an RPC node to assemble and expose the complete block, which is useful for low-latency indexing, market-data pipelines, and trading infrastructure.
What are preconfirmations? — $1,600/month
A preconfirmation, or preconf, is an early signal emitted when a Solana
leader has executed a transaction and knows its local execution result, but
before that result has been packaged into entries, converted into shreds, and
propagated through Turbine. It can therefore arrive earlier than a shred-derived
stream and earlier than the RPC processed, confirmed, or finalized
commitment levels.
Preconfirmations are valuable when a system needs to react to a likely landing or failure as soon as the leader knows the outcome. They are not the same as cluster finality: clients must reconcile the signal with the later shred, processed, confirmed, or finalized record, and availability depends on leader coverage. Exact fields, coverage, and delivery semantics are documented during onboarding.
Choosing a service
- Start with basic gRPC for conventional real-time indexing and monitoring.
- Choose very-fast gRPC when latency matters but the ultra-low-latency tier is unnecessary.
- Choose ultra-low-latency gRPC when approximately 1 ms over the $700 tier is material to the strategy.
- Choose dedicated custom gRPC for the fastest stack with isolated capacity and customer-specific tuning.
- Add shreds when you need raw block propagation before RPC assembly.
- Add preconfirmations when transaction execution status must arrive before shreds and you can correctly reconcile an early, non-final signal.
Get access
Contact @inventandchill on Telegram to discuss filters, expected throughput, deployment region, authentication, and onboarding.