Feat/OpenAI embeddings#689

Open

jonesj38 wants to merge 14 commits into

jonesj38:feat/openai-embeddings

Commits on Apr 12, 2026

feat: add OpenAI embedding support

jonesj38
authored and
Jake Jones
committed
feat: Add OpenAI embedding and query expansion support

jonesj38
authored and
Jake Jones
committed
feat: add OpenAI-based reranking via gpt-4o-mini

jonesj38
authored and
Jake Jones
committed
feat: multi-collection search support

jonesj38
authored and
Jake Jones
committed
fix: use default embedding LLM for hybrid vector queries
Jake Jones
committed
feat: split base URLs for embedding and chat endpoints

Jake Jones
and
claude
committed
Split base, chat, and rerank into separate model/url definitions so different models can be used to each. Thanks to @Kaspre for their comment
Jake Jones
committed
feat: address PR #116 feedback — base_url, expansion_model, env rename, embed fix
jonesj38
committed
feat: lazy-load node-llama-cpp, input truncation, OpenAI mode improvements

jonesj38
and
alexleach
committed

Commits on Apr 25, 2026

Merge branch 'main' into feat/openai-embeddings
jonesj38
authored

Commits on Apr 27, 2026

Commits on May 28, 2026

Merge upstream main into OpenAI embeddings branch
jonesj38
committed

Commits on Jun 2, 2026

embedding: prefer OpenAI defaults when configured
Jake Jones
committed