server-v0.0.5

anomalyco/opentuiserver-v0.0.5Jul 10, 2026by Dhravya

AI Summary

Introduces pluggable embedding model support for self-hosted deployments, allowing users to choose between local ONNX and remote providers.

Key Highlights

Support for local ONNX, OpenAI, OpenAI-compatible, and Google embedding models.
Safe upgrade checks that fail fast if model dimensions do not match to prevent vector mixing.
First-boot picker and plan locking system to manage embedding configurations.

Breaking Changes

Safe upgrades now fail fast instead of attempting to mix vectors when model dimensions differ.

New Features

Pluggable embedding model architecture.
Embedding plan locking system via 'embedding-plan.json'.
First-boot embedding model selection wizard.

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
```