Ethereum (ETH)
Ethereum is the base chain of the EVM ecosystem, with 12-second slots and finality after two epochs, about 13 minutes. Tessellate plans Ethereum RPC for launch in Q1 2027, served from Reth and Geth node pools with a Reth archive tier.
Talk to Us About Ethereum RPC
Block time
Availability target
Execution client
WebSocket subscriptions
Snapshot cadence
Head lag target
Reorg window
Debug and trace methods
*Targets are measured at our edge, per region, each month, and published on the status page. Contact our team to discuss committed capacity.
Why connect to Ethereum with Tessellate?
-
Client diversity: Reth with Lighthouse, and Geth with Prysm
-
Head lag targeted within two blocks of canonical for 99.9% of minutes
-
A Reth archive tier on local NVMe serves history and debug_ and trace_ methods
-
p95 under 250 ms for eth_call in the nearest region (target)
-
Targets published on a public status page, including the months we miss
Explore insights from our engineers
Ethereum RPC FAQ
Send JSON-RPC requests to https://ethereum.rpc.thetessellatelabs.com with your API key. Endpoints need an API key, created in the Tessellate dashboard.. See the RPC documentation at docs.thetessellatelabs.com.
We run two client pairings, so a bug in one client does not take the pool down.
Reth with Lighthouse, and Geth with Prysm. The router selects backends by health, weighted by head lag, error rate and latency.
A Reth archive tier on local NVMe serves history and the debug_ and trace_ namespaces. Mainnet archive state is of the order of 3 TB.
eth_getLogs accepts a block range of 2,000 on Developer and Build, 10,000 on Scale, and a configurable range (default 50,000) on Dedicated and Enterprise. Result caps are 10,000, 20,000 and 100,000 logs.
Archive reads are weighted 5x on Build and 3x on Scale, and are unmetered on Dedicated. For decoded events over larger ranges, the indexing product serves SQL and GraphQL. The full table of limits is on the pricing page.
Yes, from the Build plan, at 10, 50 or 500 subscriptions per API key depending on plan. Manage keys and see usage in your Tessellate dashboard. Usage, errors and latency per method and region are shown there. The current list of subscription types is in our developer docs. the pricing page.
The cache holds immutable data only, keyed by block hash. latest and pending are never cached beyond one block time. Head lag is targeted within two blocks of canonical for 99.9% of minutes. For data that must not change, use a finalised block; on Ethereum that takes two epochs, about 13 minutes. Tessellate dashboard.
No. Wallets are out of scope. Tessellate provides RPC access and never holds customer keys or assets. Use any wallet or signing tool you already trust; it connects to the endpoint like any JSON-RPC client.
Limits apply per API key and depend on the plan. Four mechanisms keep a shared pool predictable:
- Developer allows 25 requests per second and 5 million requests a month, with a hard stop. Build allows 150 per second and Scale 800.
- Overage on Build and Scale is A$1.60 and A$1.20 per million requests.
- The router classifies each method, so archive, debug_ and trace_ calls weigh more than ordinary reads.
- Dedicated plans, from A$4,500 a month, are not metered.
To learn more, read Reorgs, finality and what “latest” really means
JSON-RPC batches are 25 requests on Developer and Build, 100 on Scale and 500 on Dedicated and Enterprise. Responses are capped at 10 MB, 25 MB and 150 MB by plan.
Ethereum finalises after two epochs, about 13 minutes. Until then a block can in principle be reorganised, so treat the latest tag as provisional. Data within the 64-block reorg window used by our indexer is served with finalised:false.
Not at launch. Private transaction relay and MEV protection are routed to a partner such as Flashbots Protect on request. Mempool methods (txpool_) are on Dedicated plans only.
Sydney, Singapore and Ashburn from Q1 2027, Frankfurt in Q2 2027, and Tokyo and Dallas in Q3 2027. Latency targets are measured in the nearest region.
Shared pools run N+2 per chain per region, with snapshots every six hours to a second region and a monthly restore drill with a target of under four hours.
Explore additional chains
Tessellate supports twelve EVM chains by month twelve. Learn how to connect.
The information on this page is for general information only. It is not legal, business, tax or investment advice. Figures described as targets are goals, not guarantees; measured results are published on status.thetessellatelabs.com.
Meet With Us
Bring reliable infrastructure to your organisation.
Tessellate respects your privacy. By submitting this form, you are acknowledging that you have read and agree to our Privacy Policy, which details how we collect and use your information.