Feature comparison¶
Percona Search for MongoDB is currently available as a technical preview.
The following table compares the current capabilities of Percona Search for MongoDB with MongoDB Community Edition, MongoDB Enterprise Advanced, and MongoDB Atlas.
| Capability | Percona Search for MongoDB | MongoDB Community | MongoDB Enterprise Advanced | MongoDB Atlas |
|---|---|---|---|---|
| Deployment model | Self-managed, or Percona Operator-managed | Self-managed (tarball, Docker, or package manager); Kubernetes Operator deployment available as a Preview feature | Self-managed with the MongoDB Kubernetes Operator (Preview feature) | Fully managed |
| Full-text search | Yes | Yes | Yes | Yes |
| Vector search | Yes | Yes | Yes | Yes |
$search, $searchMeta, $vectorSearch |
Yes | Yes | Yes | Yes |
| Manual embeddings | Yes | Yes | Yes | Yes |
| Automatic embeddings | Voyage AI, OpenAI-compatible providers | Voyage AI, OpenAI-compatible providers | Voyage AI, OpenAI-compatible providers | Voyage AI, OpenAI-compatible providers |
| Replica sets | Yes | Yes | Yes | Yes |
| Sharded clusters | Yes | Yes | Yes | Yes |
| Search index management through MongoDB commands | Yes | Yes | Yes | Yes |
| Native LLM answer generation | Integrate an external LLM | Integrate an external LLM | Integrate an external LLM | Integrate an external LLM |
| Product status | Technical Preview | Generally available | Generally available | Generally available |
Note
- Percona Search for MongoDB supports manual embeddings and automatic embeddings.
- Search indexes are created and managed using
createSearchIndex(),updateSearchIndex(), anddropSearchIndex(). - Vector search supports either embeddings generated by your application or automatic embeddings generated by
mongot.