Subnet 31: rec4ll
rec4ll is Bittensor Subnet 31, a decentralized retrieval-augmented generation (RAG) network. Its on-chain description defines it as a community-owned search engine with citations: miners run the components of a RAG pipeline, validators judge the quality of results, and incoming queries are routed to whichever participants answer best.
What rec4ll Provides
Retrieval-augmented generation answers a question by first locating relevant source material and then using a language model to write an answer grounded in that material. rec4ll turns that workflow into a rewarded service on Bittensor. The subnet’s on-chain description states that miners serve embedding models, vector search, and large language model inference, while validators independently evaluate retrieval accuracy and answer quality.
The result is meant to behave like an always-improving search engine that returns answers with citations. Because miners compete openly and the network routes queries to the top performers, the quality of the pipeline is determined by ongoing evaluation rather than fixed in advance.
Miner and Validator Roles
Miners provide the retrieval-and-generation capability: converting content into embeddings, searching a vector index for relevant passages, and producing an answer with a language model. Validators provide the evaluation layer: they independently check how accurate the retrieval was and how good the resulting answer is.
Validator assessments become weights, and those weights feed into Yuma Consensus, which converts many independent evaluations into the emissions that reward the strongest miners.
On-Chain Identity
The netuid 31 identity on the Bittensor Finney network carries the subnet name rec4ll and the description quoted above. The GitHub repository URL, project URL, and Discord fields are blank in the current on-chain identity. Live data is available on TaoStats.
Relationship to Yuma Consensus
Subnet 31 uses Yuma Consensus to convert the retrieval-quality weight vectors that validators submit into the emission shares distributed to miners and validators within the subnet each tempo. The linked documentation describes how validator weight submissions are aggregated into consensus weights for each miner registered on the subnet.
rec4ll’s on-chain identity lists only a subnet name and description; the GitHub repository, project URL, and Discord fields are blank. No public source provides a source-verified description of how validators score miners on Subnet 31 beyond the on-chain description. The Emission documentation describes how consensus weights determine each participant’s share of the subnet’s accumulated emission each tempo.
Development Stage Context
The Introduction to Bittensor describes subnet development as moving from localnet to testnet and then mainnet. For rec4ll (SN31), that sequence changes how readers should interpret retrieval-augmented generation examples and search quality evaluation outcomes.
In localnet, rec4ll-compatible miners and validators can be developed and tested in an isolated environment. Localnet RAG retrieval scores and emission outcomes do not represent production subnet performance.
On testnet, rec4ll-compatible components can be exercised in a shared, non-production network. Testnet retrieval evaluations and validator scores are separate from mainnet subnet state.
On mainnet, rec4ll (SN31) is the live production subnet where miners serve embeddings, vector search, and inference for real queries, and validators evaluate retrieval accuracy and answer quality to determine real Bittensor emissions.
The Bittensor Networks reference separates mainnet, testnet, and localnet. A retrieval quality result or emission outcome from one environment should not be read as representing production subnet performance in another environment.
Netuid 31 Identifies the Subnet On-Chain
Bittensor assigns every subnet a unique numeric identifier called a netuid, and Subnet 31 is the subnet registered at netuid 31 (Glossary: Netuid). The Understanding Subnets reference explains that each subnet runs its own incentive mechanism while sharing the same underlying Subtensor chain, so the netuid is the stable handle that distinguishes rec4ll from every other subnet.
For a reader, this means “Subnet 31” and “netuid 31” refer to the same on-chain slot. A claim about rec4ll should be tied to that netuid rather than to the registered name alone, because the name field can be changed on-chain while the netuid stays fixed.
Reader Boundary
Subnet 31 rec4ll should not be read as generic Bittensor subnet documentation, a general-purpose search engine, or proof that any single retrieved answer is correct. It names one subnet’s retrieval-augmented generation competition where miners serve embeddings, vector search, and inference and validators score retrieval and answer quality on netuid 31 (Understanding Subnets, Glossary: Netuid).