Application for Filecoin ProPGF Batch 3

Beryx: Filecoin Explorer, Public API and Traces

https://github.com/Zondax/fil-propgf3-beryx

https://beryx.io

Juan Leni, CEO

Beryx is Filecoin's public chain-data layer, built and operated by Zondax: a free public API exposing the complete, queryable history of the chain since genesis, every account, transaction, internal message, and fee plus the Beryx Explorer and the Traces service. Beryx is load-bearing infrastructure with named consumers: Coinbase consumes the Beryx public API at volumes requiring a dedicated rate-limit exception, exchange-grade access to Filecoin chain data runs through this service. Filecoin Security dashboards: the Foundation's security team is building its network security dashboards on Beryx data. FIDL relies on Beryx Traces service as a chain-data source for its work in the ecosystem. FVM and smart-contract developers debug and verify contracts against the API and traces, the alternative is every team running and hosting its own indexer. Storage providers and analysts monitor sectors, deals, and fees through it; anyone can look up anything on Filecoin through the free explorer, no account required. The public API serves approximately 5 million requests per week (Cloudflare, July 2026 excluding explorer traffic). Beryx API and Explorer both won Filecoin RetroPGF in Rounds 1 and 2: the network has twice recognized this work as a public good. Beryx's coverage approaches the depth of Lily, Filecoin's reference indexer, but as a live, free public API. If this service stops, the consumers above lose their Filecoin data source or must stand up and fund a private indexing stack of its own, multiplying cost across the ecosystem and removing the one shared, verifiable view of network activity.

Partial

The Beryx Explorer source code is published publicly. The Beryx API is free and open to use under a fair-use policy, with access tokens available to anyone with a GitHub account. The indexing backend that produces the data is operated by Zondax and is not open source. An open-source release of the backend was included in the originally submitted version of this proposal (Milestone 1); following the July 2026 review discussion, that work item was removed from scope at the panel's direction, and the backend accordingly remains proprietary. The public services themselves explorer, API, and traces remain free and publicly accessible to the entire ecosystem.

Beryx is the public, genesis-complete chain-data layer for Filecoin: a free API exposing the full history of the chain, an open-source explorer built on it, and a traces service that lets developers see exactly how transactions and smart contracts executed. This proposal funds keeping those three services running, correct, and current as Filecoin and the FVM evolve. Who benefits, concretely: - Developers: every FVM/smart-contract builder debugging against the API and traces instead of running their own indexer. - Exchanges and wallets: Coinbase consumes the public API at volumes requiring a rate-limit exception; The Ledger Live Filecoin integration reads chain data through the Beryx API, hardware-wallet custody depends on this service staying up. - The Foundation and its programs: Filecoin Security is building security dashboards on Beryx data; reviewers and analysts measure network activity through it. - Storage providers and analysts:sector, deal, datacap, and fee visibility without hosting a database. - The general public: free explorer access, no account required. Current usage: ~5M API requests/week (excluding Ledger Live and explorer traffic), Beryx's indexed coverage of chain and built-in-actor data (miner and sector lifecycle, multisig, datacap, deals, verified registry, internal messages, drand, chain economics, plus FEVM/ERC-20 and decoded contract state) approaches the depth of Lily, the Filecoin project's reference chain-state indexer but exposed as a live public API rather than a node each team has to run and a database each team has to host. The 2026 strategy names chain data as a high-value area, and Beryx is the open, public version of it. The gap it fills is comprehensive, free, verifiable observability: without it, builders fall back to running their own indexers or to less complete sources, which raises cost and lowers trust across everything built on top. Beryx runs on the node and archival infrastructure in Zondax's companion proposal; this proposal funds the data and interface layer on top of it.

36000

Indirect

Direct

Indirect

**Primary objective:** Objective 2, Strengthen Network Profitability & Cryptoeconomics direct contribution to its transparency and measurement dimension; indirect to deal volume and revenue. Objectives 1 and 3 secondary, indirect. **Output**: we keep the Beryx explorer, public API, and traces available, correct, and current with the chain. **Outcome**: developers can build and debug against reliable data and execution traces, and the ecosystem, including the Foundation and reviewers, can measure and verify network activity from a free, public source. Concretely: FVM developers ship and debug against reliable data and traces; Coinbase and other integrators read Filecoin through a free public API instead of each running private indexers; Filecoin Security builds dashboards on the same data; and the general public can measure network activity from a free, externally verifiable source. **Impact**: the 2026 strategy calls explicitly for transparent impact measurement and data-driven public-goods funding, and Beryx is the open substrate that makes that possible, the same complete chain index used to see what is actually happening on Filecoin. Our direct contribution is to that measurement and transparency goal: Beryx's outputs (a genesis-complete public API, traces, and explorer) are what activity is measured through, externally verifiable via the metrics in our table. Our contribution to the headline cryptoeconomic KPIs is indirect, since we don't generate deal volume ourselves, but the link is close, because the data we publish is what those metrics are computed and trusted from. For developers (Objectives 1 and 3), the API and traces are what a builder needs to ship and debug a contract and what a flagship client needs to audit its own activity; without reliable public data, they fall back to running their own indexers or to opaque sources, raising cost and lowering trust.

| Metric | Data Source | How it's measured | Target | | Explorer availability| Zondax status page | Monthly uptime % | ≥ 99.0% monthly *| | API availability | Zondax status page | Monthly uptime % | ≥ 99.0% monthly * | | Traces availability | Zondax status page| Monthly uptime % |≥ 99.0% monthly * | *Excluding upgrade windows

Filecoin Foundation Erin O'Connor FIDL (Jon, Marta) FILPonto, (Eva Shon) Filecoin Security Dashboards (Ken) AE Studio ( mateus) Filecoin Data Portal (David Gasquez) Filecoin Builders (sarah.thiam)

Zondax has built and operated blockchain infrastructure since 2017, including Ledger applications for 30+ networks and chain-data services across multiple L1 ecosystems; Beryx has run in production for >3 years, twice recognized by Filecoin RetroPGF. **Juan Leni:** CEO and founder, Zondax. Overall responsibility for the company and the maximum engineering authority on this work. Long-standing lead of Zondax's Filecoin engagement, including the Ledger Filecoin application and the team's protocol and integration work across multiple L1 ecosystems. **Ainhoa Aldave:** Operations and delivery. Project management, partnerships, and coordination, including communication with exchanges and providers and the day-to-day running of the grant. **Emmanuel Murano:** Engineering Manager. Architecture and engineering oversight of the Beryx indexing backend, public API, and Traces service; responsible for data correctness and operational delivery through network upgrades.

The main dependency is the companion infrastructure proposal: Beryx runs on the node and archival infrastructure funded there, so if that isn't funded, Beryx has no data to serve regardless of this grant. The two are scoped and measured separately at the funder's request but are technically linked, and this proposal's pricing assumes the companion proposal is accepted. Following the July 2026 review discussion, the open-source release of the indexing backend was removed from this proposal's scope at the panel's direction. The backend remains proprietary, and single-provider dependency on the indexing layer is accordingly not mitigated within this engagement. The main external dependency is the Filecoin and FVM upstream: protocol and FVM changes can alter data formats and execution semantics, which means the indexer, API, and traces need updating so results stay correct through each upgrade, and an unexpected change can briefly affect data currency. This is a maintenance-level scope: at this budget the project keeps the services running and correct rather than building new features, which would need separate funding. The work has no commercial revenue behind it, so a funding gap stops it rather than degrading it gradually.

This is one of two complementary Batch 3 submissions from Zondax, revised following the July 21 review discussion: this proposal was reduced by removing the one-time open-source release of the indexing backend (originally Milestone 1) per the panel's decision, leaving pure operation of the three public services. This proposal covers the data and interface layer (explorer, API, traces) that FVM developers, Coinbase, Filecoin Security, storage providers, and the general public consume; the companion proposal funds the node and archival infrastructure that Beryx and other consumers run on. Both proposals now share an aligned term.