Cito

What it does

Cito is a hybrid search engine built on top of the Semantic Scholar corpus. It indexes 236 million academic papers for keyword search and 146 million of those papers with SPECTER2 dense vectors for semantic search. When a query is issued, results from both keyword and semantic searches are combined using Reciprocal Rank Fusion (RRF) and then re-ranked by a cross-encoder model to improve relevance. The service is accessible through a free web interface that requires no signup, but its primary value lies in its programmatic access: a plain JSON API and a native MCP (Model Context Protocol) endpoint. The MCP endpoint allows AI agents—such as Claude Code—to perform deep literature searches without hitting the rate limits that plague other academic APIs. The creator explicitly built Cito because "every academic API throttled my agents to death."

Who it is for

Cito targets three main user groups: (1) AI agents and their developers who need a scalable, rate-limit-free source of academic papers for tasks like literature review, research synthesis, and retrieval-augmented generation; (2) researchers who prefer a fast, no-signup search interface for quick paper discovery; and (3) builders of custom tools who want to integrate academic search via a simple JSON API. The MCP endpoint is especially relevant for users of agent frameworks that support the Model Context Protocol, enabling seamless integration with minimal setup.

Why it matters

Academic search is critical for keeping up with the ever-growing volume of published research, but most available APIs impose rate limits that make them unsuitable for automated or agent-based use. Cito addresses this gap by offering an open, free endpoint designed from the ground up to handle agent-scale queries. The hybrid search approach improves result quality: keyword search ensures exact matches and coverage, while dense vector search captures semantic similarity, and cross-encoder reranking further refines results. This combination is particularly valuable for tasks like finding related work, benchmarking, and systematic reviews. By building on the Semantic Scholar corpus, Cito leverages a well-known academic dataset while adding its own search infrastructure.

Launch signal

Cito debuted on Product Hunt as a new product. The launch post introduces its features and motivation. The product has already garnered 45 followers on the platform. The founder, Tao An, positions Cito as a solution to the throttling problems encountered when using existing academic APIs with agents. The tagline "Hybrid academic search over 236M papers, built for agents" communicates its key differentiation. At launch, the web interface appears minimal (displaying a set of landmark papers as starting points), suggesting that the product's main functionality is accessed through its API and MCP endpoint.

Brand and naming

The name "Cito" is short and evokes speed (from Latin "citus"), aligning with the product's emphasis on efficient hybrid search. However, the name does not immediately convey the domain of academic search, and the likely unavailability of cito.com may limit brand recall. The tagline and Product Hunt description clearly position it as "built for agents," which differentiates it from general academic search engines like Google Scholar or Semantic Scholar itself. The brand is still nascent, with limited public awareness beyond the Product Hunt community.