Application for Filecoin ProPGF Batch 3

IPNI: Interplanetary Network Indexer

https://www.github.com/ipni

https://www.cid.contact

Alex Firmani, General Manager

IPNI is the content routing layer for the Filecoin and IPFS ecosystems. It answers “who has this piece of content” by maintaining a global index that maps content identifiers (CIDs) to storage providers (SPs). For any use of Filecoin as a retrieval network, a client must first discover which provider holds the requested content. The DHT does not scale to the content footprint of large Filecoin providers, so those providers, and the integrations built on top of them, depend on IPNI. Direct dependents today include Spark (retrieval checking), IPFS gateway operators, Storacha, Curio, and on-ramps bridging IPFS and Filecoin.

Fully Open Source

N/A

The InterPlanetary Network Indexer (IPNI) is the content routing layer for the Filecoin and IPFS ecosystems. It maintains a global index mapping content identifiers to the storage providers that hold them, and answers low-latency lookup queries from retrieval clients, gateways, and on-ramps. Storage providers publish signed advertisements of their content; IPNI ingests those advertisements and serves billions of lookups per month. This grant funds six months of continued operation and forward development of IPNI, structured around a single strategic objective: making the paid publisher tier deliverable and sellable. The prior ProPGF Financial Federation Design grant (batch 2) concluded that paid tiers are the credible path to IPNI's financial sustainability. This batch funds the engineering work that turns that design into a real capability and revenue streams. The engineering scope is the Ingest v2 workstream: sync-on-announce scheduling plus a dedicated paid-tier ingest lane on existing unused hardware (M1), followed by cache invalidation infrastructure at the edge for paid-tier publishers (M2). Together these two milestones deliver end-to-end web-speed ingest latency (target: one to three seconds from publish to queryable) as a contractually-measurable capability for paying customers. The new paid tier is an additive capability built on engineering investment, not a paywall on previously free features.

$280,000

Indirect

Indirect

Indirect

Not Provided

Not Provided

Not Provided

Not Provided

The most material risk to this grant's outcomes is publisher adoption of the paid tier. The value of that capability depends on at least a few large publishers signing on as paying customers. Single-engineer dependency is a structural risk. IPNI's engineering roadmap currently runs through one full-time software engineer; the project has limited upkeep time designated to the storetheindex codebase and related libraries. The infrastructure operational model carries structural debt. The current setup depends on bare-metal hardware co-located in a rack sub-leased through the technical advisor, with the advisor providing physical labor for hardware maintenance. This arrangement is functional but is not how IPNI's infrastructure would be built today, and creates a single-point dependency on one advisor's continued availability. The project intends to evaluate restructuring options (direct datacenter lease, managed colo, or selective cloud migration) in a future grant cycle. This grant funds continued operation of the current arrangement at fair market advisory rates while that evaluation is scoped.

Not provided

Not provided