Onchain events in your Kafka within seconds
Full-chain streams from 80+ chains at 3-5s, or Beam filtered streams at P50 under 2s. Delivered to Kafka, Pub/Sub or SNS.
Why teams choose Allium over webhooks and subgraphs
Webhooks
Best-effort delivery means messages are lost during outages or system downtime.
Subgraphs
Raw blockchain events require specialized knowledge to decode and query.
GraphQL adds complexity and learning curve.
Indexing costs scale exponentially with query volume, making production usage expensive.
Teams need blockchain expertise to maintain and troubleshoot.
Guaranteed delivery with enterprise message brokers (Kafka, Pub/Sub, SNS)
Decoded & enriched data with consistent schemas
Standard SQL without graph expertise
Predictable scaling costs for production workloads
Broader data platform: streaming, analytics, compliance from one source
Real-time, production-ready streams
Push-based data streams for wallets, tokens, and trades.
Choose
Filter
Deliver
Reconcile
.avif.avif)
.avif.avif)
Choose the streaming setup that best fits your infrastructure
Datastreams
Allium Beam
Beta
What it is
Datastreams
Full blockchain data streams delivered to your message broker
Allium Beam
Beta
Custom-filtered stream transformations — define filters in a UI, receive only what you need
Best for
Datastreams
Teams with streaming infrastructure who want complete chain data
Allium Beam
Beta
Teams who want filtered streams without building or maintaining infrastructure
Data
Datastreams
Full chain topics (e.g., ethereum.transactions, solana.dex_trades)
Allium Beam
Beta
Custom-filtered + transformed events
Latency
Datastreams
P50 ~5 seconds
Allium Beam
Beta
P50 <2 seconds
You handle
Datastreams
Filtering, transformation, routing
Allium Beam
Beta
Just the business logic
Sinks
Datastreams
Kafka, Pub/Sub, SNS
Allium Beam
Beta
Kafka, SNS, Pub/Sub, direct DB (ClickHouse, Postgres)
Setup time
Datastreams
Hours
Allium Beam
Beta
30 seconds
Engineering effort
Datastreams
Moderate
Allium Beam
Beta
Minimal
Datastreams
Subscribe to full chain topics and process events in your infrastructure.
- Choose topics: ethereum.transactions, solana.dex_trades, etc.
- Connect your Kafka or Pub/Sub: events flow automatically
- Build your processing: filter, transform, route to your systems
Allium Beam
Beta
Define filters in a UI and receive only what you need
- Define your filters: Contract addresses, wallet set lookups (100K+ addresses), value thresholds
- Add transformations: Reshape data, extract fields, enrich events
- Choose your sink: Kafka, SNS, Pub/Sub, direct DB
- Deploy in 30 seconds: Visual builder or AI Beam Builder
Supported integrations
Kafka
Kafka
Apache Kafka Integration
Pub/Sub
Pub/Sub
Google Cloud Pub/Sub Integration
SNS
SNS
Amazon SNS Integration
ClickHouse
ClickHouse
Direct database writes (Beam)
Postgres
Postgres
Direct database writes (Beam)
Websockets
Websockets
Get APIs or datastreams over websockets
Phantom, MetaMask and Magic Eden run real-time features on Allium
Built on Allium's dedicated Hyperliquid node, our realtime Hyperliquid dashboard surfaces trades, validator activity, and protocol metrics as they happen onchain.
Magic Eden relies on Allium to filter inorganic marketplace activity and power real-time transaction data in their wallet, for accurate market intelligence and user insights.
Phantom uses Allium Developer APIs to power wallet balances and transaction histories for millions of users. During Jupiter's 2024 airdrop, Allium handled 90,000 requests per second with zero downtime.
MetaMask uses Allium APIs to power real-time balance calculations and transaction data across their wallet infrastructure.
FAQ
Should I use Datastreams or Beam?
Use Datastreams if you have streaming infrastructure and want full chain data with maximum control. Use Beam if you want custom-filtered stream transformations without building or maintaining real-time infrastructure, ideal for notifications, monitoring, and alerting.
What's the latency difference?
Datastreams delivers at P50 ~5 seconds. Beam delivers at P50 under 2 seconds. Beam is faster because filtering and transformation happen at source through co-located processing.
Can I use both?
Yes. Some teams use Datastreams for analytics (full data to data lake) and Beam for user-facing features (custom-filtered streams for notifications).
What kind of filters and transformations does Beam support?
Filters: contract address filtering, large wallet set lookups (100K+ addresses), value thresholds, event type selection. Transformations: data reshaping, field extraction, event enrichment, all defined in the UI.
Does Beam support my sink?
Beam supports Kafka, AWS SNS, Google Pub/Sub, and direct database sinks (ClickHouse, Postgres).
How much engineering time does Beam save?
Building custom streaming pipelines typically takes 3-6 months. Beam deploys in 30 seconds with a visual builder or AI Beam Builder.
How much infrastructure does Beam eliminate?
Beam eliminates downstream Kafka consumers, transformation pipelines, and filtering logic.
Why Allium over Goldsky?
Goldsky Mirror is streaming-only. Allium gives you streaming plus Explorer, Datashares, APIs, and attribution data, all from the same standardized schemas. When multiple teams reference the same Allium data, switching creates coordination cost across finance, product, engineering, and compliance.
1. Start free
With 20,000 credits.
2. Explore
In Terminal or SQL across 10,000+ protocols.
3. Deliver
The same data to your warehouse, APIs or streams.
Stream onchain data to your stack
Start free with 20,000 credits. Paid plans from $200/month.
