Build with Membase
Start with one working API call sequence, then learn memory operations, integrate your model, and look up SDK and API details.
Use Membase from code to add material to a person’s memory and retrieve it for your app.
Calls use that person’s developer key and the Memories they grant it. Python, TypeScript
and REST all reach the same service at https://api.app.membase.io.
Start with the quickstart. It creates one document, waits for learning to finish, and searches the result. The example includes the checks needed to distinguish an empty Memory from a failed search.
Choose your next task
Section titled “Choose your next task”| Task | Guide |
|---|---|
| Add facts or documents, search, read the profile, forget or delete | Memory operations |
| Serve multiple people | Multi-user isolation |
| Configure timeouts, retries or SDK methods | SDKs |
| Set access and reach, rotate or revoke a credential | Authentication |
| Look up an endpoint or response shape | API reference |
| Fix a failed request or unread document | API troubleshooting |
Integrations
Section titled “Integrations”| Your application | Guide |
|---|---|
| Claude API | Claude API |
| OpenAI API or a function-calling model | OpenAI API |
| An agent framework that speaks MCP | MCP frameworks |
| A coding assistant building the integration for you | AI coding assistants |
If you want to give an existing AI app your memory without building an integration, use Connect your AI.
Understand the model
Section titled “Understand the model”In the API, a container is one Memory in the app, a document is raw material it reads, and a memory is a learned fact or a profile fact. A container is not an end user.
Platform overview maps these objects to the app. How Membase works explains learning, retrieval and model requirements. The engine benchmark report is separate from API setup and performance guidance.