Summary: DipDup is a Python framework for building smart-contract indexers and application backends with a strong emphasis on selective ingestion, developer-defined schemas, and operational tooling around the resulting API surface. Its core mechanism is not a decentralized query network but a programmable indexing framework that lets teams specify exactly which contracts, events, and data flows matter, then expose the results through GraphQL, REST, and database-backed services. That makes it a useful comparison class for The Graph, SubQuery, SQD, and hosted indexing platforms because it shifts practical control toward application-specific indexer design rather than shared network-wide data availability.
What it does:
Lets developers build selective smart-contract indexers in Python so only required chain data is requested and stored
Supports multiple ecosystems including EVM, Starknet, Tezos, and Substrate through chain-specific quickstarts and index types
Bundles operational surfaces beyond indexing itself, including GraphQL, REST endpoints, database integrations, monitoring, backups, Prometheus, Sentry, and reindexing workflows
Uses YAML configuration plus Python packages and handlers to define contracts, indexes, models, hooks, jobs, and API behavior
Extends beyond the main framework through related tooling such as DipDup’s modular Go indexer-sdk, which presents indexing as flow-based pipelines of sources, processors, and sinks
Key claims:
The official README and docs describe DipDup as a Python framework for building smart-contract indexers that reduces boilerplate and lets developers focus on business logic
DipDup’s primary differentiator in first-party materials is selectivity: its README says DipDup-based indexers request only required data, aiming for faster indexing and lower load on upstream APIs
The docs navigation shows the project is broader than a minimal indexing SDK, spanning config-driven indexes, GraphQL and REST serving, hooks, jobs, database engines, monitoring, and deployment guidance
The public repo says the framework is maintained by the Baking Bad team, with development supported by the Tezos Foundation and OnlyDust, which is relevant context for how the project matured from Tezos roots into broader multi-chain indexing infrastructure
DipDup’s separate indexer-sdk repo describes a flow-based programming approach with modular sources, processes, and sinks, which suggests an additional path toward more composable indexing pipelines beyond the main Python framework
Whitepaper: No classic whitepaper or litepaper was found during this pass. The strongest primary materials were the official docs portal, the public GitHub repositories, and first-party READMEs; see ../whitepapers/dipdup-primary-sources-2026-05-09.md.