> Documentation index: https://rift.sh/llms.txt

# Rift vs Zep

Zep builds graphs of dated facts from conversations and business data and serves them from Zep Cloud. Rift syncs files between machines.

### What Zep is

Zep is a context service for agents. It takes chat messages, JSON and text, and builds a graph for each user, and shared graphs for context that no one user owns, in which every fact has dates and is marked invalid when newer data contradicts it. An agent gets the context relevant to its task in one call, and each fact can be traced to the source it came from. Enterprise plans add policies that limit which actions and which classes of data each API key can reach. Zep is a managed cloud service, and its graph framework, Graphiti, is open source.

### How they differ

Zep stores facts and relationships that it extracts from what you send, in graphs on Zep Cloud, where its services decrypt the data to process it. Rift stores files as they were written and shares them one folder at a time with keys that the receiving machines hold.

- Identity
  - Zep: A Zep project API key.
  - Rift: A key that each machine and agent makes for itself. No account.

- Smallest share
  - Zep: A project API key. On the Enterprise plan a policy can limit the key to some actions and some classes of data.
  - Rift: A folder, a subfolder or one file, as read or write.

- Who checks access
  - Zep: Zep's servers, on each request.
  - Rift: Each machine that receives a file checks the grant itself.

- Who can read your files
  - Zep: Zep. Its documentation says automated services decrypt data to process it. An Enterprise add-on keeps the encryption key in your own AWS account, and an Enterprise deployment can run inside your own VPC.
  - Rift: Only the machines you shared with. Rift Cloud holds ciphertext.

- When your machines are off
  - Zep: Graphs stay available from Zep Cloud.
  - Rift: Files stored in Rift Cloud stay available.

### Which to choose

Choose Zep when

- Agents need facts that change over time, with a record of when each was true.
- Several agents should draw on one body of business context.
- You need policies on what each agent can retrieve, and logs of its requests.

Choose Rift when

- The state is files, such as reports, datasets and video.
- The files have to be on the agent's own disk.
- A cloud should store the files and be unable to read them.

### Using them together

An agent can keep facts in Zep and files in Rift.

[Use Rift with Zep](https://rift.sh/docs/use-with/zep.md)

Based on Zep's documentation, read in October 2026. Sources: [concepts](https://help.getzep.com/concepts.md) , [agent access](https://help.getzep.com/attribute-based-access-control.md) , [bring your own key](https://help.getzep.com/bring-your-own-key.md) , [support access](https://help.getzep.com/support-access.md) , [security](https://help.getzep.com/security-compliance.md) , [Zep and Graphiti](https://help.getzep.com/zep-vs-graphiti.md).