HN Explorer

Show HN: OctaneDB – Fast, Open-Source Vector Database for Python (github.com)

33 points by rijin_r ↗ HN
OctaneDB is an open-source vector database for Python that focuses on ultra-fast similarity search for high-dimensional data—perfect for AI/ML, semantic search, and large-scale document or embedding retrieval.

What does it do?

Store, index, and search millions of embeddings (text, images, etc.) with sub-millisecond query time.

Supports in-memory and efficient HDF5 persistent storage.

Integrates seamlessly with sentence-transformers for automatic text embedding.

Key Features:

10x faster than Pinecone or ChromaDB for vector search and batch insertions.

Advanced indexing: HNSW (approximate nearest neighbor), FlatIndex

Batch search, advanced metadata filtering, GPU acceleration

8 comments

[ 3.4 ms ] story [ 21.2 ms ] thread
ozgrakkurt
Please stop with these LLM generated readme files. It burns the eyes.

Just take 10 minutes to write something that explains the project.

As soon as I see this README I assume the code is generated(worthless) as well

Bishonen88
Looking at the repo, everything seems too clean and pristine. There's only 5 commits. Not that it matters, but was the whole thing vibe coded? If so, I wonder if mentioning this in the Readme would be helpful to potential users.
JSR_FDED
This looks very good, easy to understand. Do you have a sense for how much RAM it uses (not for storing the vectors themselves - those I’ll keep in a file), but when doing a search? I have one use case where it could run on a VPS provided its memory use doesn’t balloon too much.
icemanx
Definitely vibe code, and I agree with the others - please mention it in the README if you want people to use AI generated code. You are not the mind behind it

Update: Looking through your code, I already found within 5 minute flaws, since it is obvious that you have not written it (looking at your past work, I doubt you even understand what it does), I will not even point the issues out.

brandonasuncion
Claims of being "10x faster performance than existing solutions" using a HNSW written entirely in Python is enough to raise alarm bells, vibecoded or not.
rf15
Looking at your code, I don't think you understand any of the regular DB requirements, let alone how to implement a vector database that is actually fast. Because, let me tell you: this isn't it.

Please get more experience and use AI less. It's not a good look, and if it's vibe coded it's not your code anyway, it's just you asking somebody else to do it (at best). You don't want to be an expert in asking others to do your work.

Best of luck on your turn around.

dikei
Big claims with not a single benchmark to back them up
redskyluan
Curious—why the shift from a Milvus-compatible API to a Chroma-compatible one? And of course, something in Python… because that’s obviously the fastest way to conquer the world.