# Into: Recall A subnet that turns retrieval-augmented generation into an open contest: nodes compete to serve the best search-and-answer pipeline, and validators score who actually retrieves the right thing. // Community-owned RAG, ranked by competition --- > New to Bittensor? Start here. Experienced users can skip to the full analysis. ### What is Recall? Recall is a Bittensor subnet (SN31) built around retrieval-augmented generation, or RAG: the technique of answering a question by first fetching relevant documents and then generating a response grounded in them, with citations. According to its on-chain identity description, miners serve embedding models, vector search, and LLM inference, while validators independently evaluate retrieval accuracy and answer quality. **The simple version:** It's like Perplexity, but the search-and-answer engine is run by a competing network of nodes instead of a single company, and the best pipeline wins the queries. **Centralized equivalent:** Perplexity, or any managed RAG stack that bolts a vector database onto an LLM. **How it works:** - **Miners** serve the retrieval and generation pipeline: embedding models, vector search, and LLM inference, per the subnet's on-chain description. - **Validators** check retrieval accuracy and answer quality, scoring which pipelines actually surface and use the right information. --- ### Why This Matters - **The problem it solves:** Language models answer from frozen training data and confidently invent things they do not know. RAG grounds answers in freshly retrieved sources with citations. Recall's stated goal is to make that retrieval-and-generation pipeline an open competition rather than a closed product. - **The opportunity:** Retrieval is the backbone of most useful AI systems today. A continuously improving, citation-backed engine that anyone can plug into is a large surface to aim at. - **The Bittensor advantage:** Open competition routes user queries to the top-performing pipeline and keeps pressure on every node to improve, instead of locking users to one vendor's retrieval quality. - **Traction signals:** On-chain, the alpha token has positive net inflows over the past week and rose roughly 26% in that window. Beyond the on-chain identity description and market data, the slot does not list a public website, repository, or social account, so independent verification of the implementation is limited for now. --- ## Full Analysis **Category:** Search and Information Retrieval | **Centralized Competitor:** Perplexity, managed RAG stacks Retrieval-augmented generation has become the default pattern for making language models useful on private or current data: retrieve relevant context, then generate grounded answers. Recall frames this as a subnet-level competition, where the question is not which company built the best RAG stack but which node on the network does, measured live. **Mechanism:** Per the subnet's on-chain identity description, miners run the full pipeline: embedding models to represent queries and documents, vector search to find candidates, and LLM inference to produce the answer. Validators independently evaluate retrieval accuracy and answer quality, and the subnet routes queries toward the pipelines that score best. In Bittensor terms, validators set weights based on those quality evaluations, and emissions follow the weights, so the strongest retrieval-and-answer pipeline earns the most. On market footing, the alpha token trades near 0.00476 TAO with a market value around 7,065 TAO. Its emission share sits near half a percent of network emissions, backed by roughly 1,773 TAO of root TAO in the pool. Net inflows over the past week have been positive and price is up about 26% on the week, though it is roughly flat to slightly lower over the past month. The pool is on the thinner side, so position sizing matters. A note on sourcing: the mechanism above comes entirely from the subnet's on-chain identity description, which is the authoritative current claim from the slot owner. A separate one-line summary mirrored in a third-party database still describes an older, unrelated theme; that is stale data and the on-chain identity is what governs here. --- ### Risk Factors - **Limited public verification:** The on-chain identity does not register a website, repository, or documentation, so the mechanism is known only from a short on-chain description. Independent review of the actual implementation is not currently possible from public sources. - **Crowded category:** Retrieval and search is contested both inside Bittensor, where several subnets target search and information retrieval, and outside it, where well-funded products like Perplexity already serve RAG at scale. - **Concentration and liquidity:** A Gini coefficient near 0.76 points to concentrated ownership or stake distribution, and with only about 1,773 TAO of root TAO backing the pool, large positions can move price and incur slippage on entry or exit. - **Early execution:** This is an early-stage subnet whose public footprint is mostly its on-chain identity and market data. Execution against the RAG goal is unproven from what is publicly verifiable. --- Into the next one.