--- title: "hora vs cherche" type: "comparison" canonical_url: "https://www.graphcanon.com/compare/hora-search-hora-vs-raphaelsty-cherche" tools: ["hora-search-hora", "raphaelsty-cherche"] --- # hora vs cherche *GraphCanon updated Aug 23, 2026* ## Verdict Pick hora if hora is a Rust-based library for high-performance approximate nearest neighbor search algorithms critical in applications such as image and recommender systems where speed and efficiency are paramount; pick cherche if cherche is a Python library for implementing neural search capabilities. [hora](http://horasearch.com/) reports 2.7k GitHub stars, 77 forks, and 26 open issues, last pushed Feb 17, 2026. [cherche](https://github.com/raphaelsty/cherche) has 332 stars, 14 forks, and 4 open issues, last pushed Jun 1, 2024. Figures are from public GitHub metadata via [hora's repository](https://github.com/hora-search/hora) and [cherche's repository](https://github.com/raphaelsty/cherche). | | [hora](/tools/hora-search-hora.md) | [cherche](/tools/raphaelsty-cherche.md) | | --- | --- | --- | | Tagline | efficient approximate nearest neighbor search algorithm collections library written in Rust | Neural Search | | Stars | 2,659 | 332 | | Forks | 77 | 14 | | Open issues | 26 | 4 | | Language | Rust | Python | | Adopt for | Hora is a Rust-based library for high-performance approximate nearest neighbor search algorithms critical in applications such as image and recommender systems where speed and efficiency are paramount. | Cherche is a Python library for implementing neural search capabilities. | | Persona | - | - | | Runtime | - | - | | License | Apache-2.0 | MIT | | Categories | Data & Retrieval, Vector Databases | Data & Retrieval, Evaluation & Observability, Vector Databases | ## Trust and health _Sourced signals - not a safety guarantee. No winner column._ | | [hora](/tools/hora-search-hora.md) | [cherche](/tools/raphaelsty-cherche.md) | | --- | --- | --- | | Maintenance | Slowing (36%) | Dormant (18%) | | Days since push | 186d | 812d | | Open issues (now) | 26 | 4 | | Stars delta | -4 (30d) | 0 (30d) | | Owner type | Organization | User | | Full report | [trust report](/tools/hora-search-hora/trust.md) | [trust report](/tools/raphaelsty-cherche/trust.md) | ## Shared compatibility - **Python**: [hora](/tools/hora-search-hora.md) - Python runtime; [cherche](/tools/raphaelsty-cherche.md) - Python runtime ## Decision facts: hora - **Adopt for:** Hora is a Rust-based library for high-performance approximate nearest neighbor search algorithms critical in applications such as image and recommender systems where speed and efficiency are paramount. ## Decision facts: cherche - **Adopt for:** Cherche is a Python library for implementing neural search capabilities. ## Choose when ### Choose hora if… - hora is primarily Rust; cherche is Python. - License: hora is Apache-2.0, cherche is MIT. - Tags unique to hora: algorithm, approximate-nearest-neighbor-search, artificial-intelligence, data-structures. - When you require a high-speed library written in Rust to implement an efficient similarity-search solution ### Choose cherche if… - cherche is primarily Python; hora is Rust. - License: cherche is MIT, hora is Apache-2.0. - Tags unique to cherche: bm25, flashtext, information-retrieval, machine-learning. - Also covers Evaluation & Observability. - Cherche is a Python library for implementing neural search capabilities. ## When NOT to use hora - If your project is restricted to languages other than Rust and you cannot work with Rust bindings or require a more general-purpose vector database library - When exact nearest neighbor search accuracy is preferred over the speed provided by approximate algorithms, as Hora specializes in sacrificing minimal precision for significant performance gains ## When NOT to use cherche - Last GitHub push was 815 days ago (dormant maintenance, Jun 1, 2024). Validate activity before betting a new project on cherche. - Data & Retrieval: Skip a heavy ingestion framework when your corpus is small and static; a script plus the embedding API is enough. - Evaluation & Observability: Defer heavyweight eval infra only until you have real traffic - never skip it once users depend on answers. - Vector Databases: Don't reach for a dedicated vector DB under ~100k vectors; pgvector on your existing Postgres is simpler to operate. ## Common questions ### What is the difference between hora and cherche? hora: efficient approximate nearest neighbor search algorithm collections library written in Rust. cherche: Neural Search. See the comparison table for live GitHub stats and shared categories. ### When should I choose hora over cherche? Choose hora over cherche when hora is primarily Rust; cherche is Python; License: hora is Apache-2.0, cherche is MIT; Tags unique to hora: algorithm, approximate-nearest-neighbor-search, artificial-intelligence, data-structures; When you require a high-speed library written in Rust to implement an efficient similarity-search solution. ### When should I choose cherche over hora? Choose cherche over hora when cherche is primarily Python; hora is Rust; License: cherche is MIT, hora is Apache-2.0; Tags unique to cherche: bm25, flashtext, information-retrieval, machine-learning; Also covers Evaluation & Observability; Cherche is a Python library for implementing neural search capabilities. ### When should I avoid hora? If your project is restricted to languages other than Rust and you cannot work with Rust bindings or require a more general-purpose vector database library When exact nearest neighbor search accuracy is preferred over the speed provided by approximate algorithms, as Hora specializes in sacrificing minimal precision for significant performance gains ### When should I avoid cherche? Last GitHub push was 815 days ago (dormant maintenance, Jun 1, 2024). Validate activity before betting a new project on cherche. Data & Retrieval: Skip a heavy ingestion framework when your corpus is small and static; a script plus the embedding API is enough. Evaluation & Observability: Defer heavyweight eval infra only until you have real traffic - never skip it once users depend on answers. Vector Databases: Don't reach for a dedicated vector DB under ~100k vectors; pgvector on your existing Postgres is simpler to operate. ### Is hora or cherche more popular on GitHub? hora has more GitHub stars (2,659 vs 332). Stars measure visibility, not whether either tool fits your constraints. ### Are hora and cherche open source? Yes - both are open-source projects on GitHub (hora: Apache-2.0, cherche: MIT). ### Where can I find alternatives to hora or cherche? GraphCanon lists graph-backed alternatives at [hora alternatives](/tools/hora-search-hora/alternatives) and [cherche alternatives](/tools/raphaelsty-cherche/alternatives) ([hora markdown twin](/tools/hora-search-hora/alternatives.md), [cherche markdown twin](/tools/raphaelsty-cherche/alternatives.md)), ranked by typed relationship edges rather than popularity votes. ### Is there a machine-readable version of this comparison? Yes. The markdown twin at [this comparison](/compare/hora-search-hora-vs-raphaelsty-cherche.md) mirrors this page for agents and LLM crawlers, with the same stats table and FAQ answers. ### Which is better maintained, hora or cherche? hora: Slowing. cherche: Dormant. Compare maintenance labels, days since push, and release cadence in the trust section below - stars alone do not measure maintenance. ### Where are the full trust reports for hora and cherche? GraphCanon publishes per-repo trust reports with dated maintenance, provenance, and scan summaries: [hora trust report](/tools/hora-search-hora/trust); [cherche trust report](/tools/raphaelsty-cherche/trust). --- **Machine-readable endpoints** - JSON: [`/api/graphcanon/graph?tool=hora-search-hora`](/api/graphcanon/graph?tool=hora-search-hora) - LLM index: [/llms.txt](/llms.txt) - Full corpus: [/llms-full.txt](/llms-full.txt) _GraphCanon - The knowledge graph for AI development. https://www.graphcanon.com/_