server-v0.0.5

achtungsoftware/alarikserver-v0.0.5Jul 10, 2026by Dhravya

AI Summary

This release introduces a pluggable embedding architecture for self-hosted instances, allowing the use of local ONNX models or remote providers like OpenAI and Google.

Key Highlights

Pluggable embeddings support for local ONNX (default) and remote OpenAI/Google models.
Configuration via `SUPERMEMORY_EMBEDDING_*` environment variables.
Safe upgrade logic prevents vector mixing by failing fast on dimension mismatches.
First-boot picker and plan locking via `embedding-plan.json`.

New Features

Pluggable embedding models
Environment variable configuration for embeddings
Safe upgrade mechanisms

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
```