server-v0.0.5

opensandbox-group/OpenSandboxserver-v0.0.5Jul 10, 2026by Dhravya

AI Summary

Self-hosted supermemory now supports pluggable embedding models, allowing users to choose between local ONNX or remote providers like OpenAI and Google. The update includes a first-boot picker and plan locking mechanism to manage embedding configurations securely and safely upgrade existing instances.

Key Highlights

Pluggable embeddings support (ONNX, OpenAI, OpenAI-compatible, Google)
Configuration via environment variables (SUPERMEMORY_EMBEDDING_*)
First-boot picker and plan locking in embedding-plan.json
Safe upgrades with fail-fast behavior for dimension mismatches
Legacy model support for populated stores without a lock

Breaking Changes

Model dimension mismatches now fail fast instead of mixing vectors
Embedding plan locking introduced which may affect existing configurations

New Features

Local ONNX embedding support
Remote embedding integration (OpenAI, OpenAI-compatible, Google)
First-boot embedding model picker
Plan locking for embedding configuration
Safe upgrade logic to prevent vector mixing

Full Release Notes

Self-hosted supermemory can now use local or remote embedding models.

- **Pluggable embeddings** — local ONNX (default) or OpenAI / OpenAI-compatible / Google via `SUPERMEMORY_EMBEDDING_*`; first-boot picker + plan lock in `embedding-plan.json`
- **Safe upgrades** — populated stores without a lock assume legacy `local · Xenova/bge-base-en-v1.5 · 768d`; same-dimension model switches fail fast instead of mixing vectors

```sh
curl -fsSL https://supermemory.ai/install | bash
```

```sh
supermemory-server upgrade
```