Top all-time · #9772
Mixedbread
Company blog · mixedbread.com/blogaidata
Mixedbread is the company blog at mixedbread.com, writing about ai and data. Since 2025, 8 different Hacker News readers have linked it 23 times, most often “Dense Retrievers Know More Than They Can Express”. It is still writing: the last post was on 2026-08-14, and you can follow it by feed.
Too few sampled comments (2) to say how often HN recommends it.?
Why HN links it
HN readers link Mixedbread for its retrieval and search work, including its open-sourced models and blog posts. Comments point to the blog as a source of learning and of open-source releases.
Summarised by Claude from a sample of 2 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
Make your agents work on your data. Mixedbread turns PDFs, decks, videos, code, Slack, and Drive into evidence your agent can act on, with fewer tokens.
Most-linked posts?
- Dense Retrievers Know More Than They Can Express2 people · 2 mentions · first linked 2026-06-02
- Introducing Toast 12 people · 2 mentions · first linked 2026-08-13
- Building a Robust Ingestion System for Any File of Any Size2 people · 2 mentions · first linked 2026-07-15
What Hacker News says?
Thats actually not correct. Embeddings can handle relationships like “without” or “not.” when trained for it. You need to scale up the training massively to make it generalize it well. The current version of Mixedbread Search supports negatives like "tshirt without stripes". You can check it out on our launch video [1]. We are working on a way more generalized model, which should be able to capture relationships, em…
Comment linking /blog/mixedbread-search
on “Show HN: Semantic search over the National Gallery of Art”Submitted Introducing Toast 1 as a story · 39 points
HN discussionyou can look it up in the blog. RAG is not super great because of two reasons, single embedding vector models are not that good and stopped improving and second most models are not good at looking up information. we spend great time on improving the modeling side by inventing on the indexing level [1, 2]. and now we trained our model to be very good at search. it is matching the quality of Opus 5 and GPT 5.6 Sol whi…
Comment linking /blog/wholembed-v3
on “Introducing Toast 1”Submitted Maxsim-CPU: Maximising Maxsim Efficiency as a story · 3 points
HN discussionyou can plugin your existing stack and use it via an openai compatible client. https://www.mixedbread.com/docs/agent/chat-completions
Comment linking /docs/agent/chat-completions
on “Introducing Toast 1”Submitted Mxbai-Rerank-v3.1-Listwise as a story · 2 points
HN discussionMixedbread Search is a multimodal & multilingual search product, where you can upload any kind of data and make it searchable. Its powered by Wholembed [1] v3, a late interaction retrieval model. [1]: https://www.mixedbread.com/blog/wholembed-v3
Comment linking /blog/wholembed-v3
on “Introducing Toast 1”Submitted Dense Retrievers Know More Than They Can Express as a story · 2 points
HN discussion*MixedBread https://www.mixedbread.com/
Comment linking /
on “Beating GPT-5.6 Sol on retrieval with 100x cheaper open models”Submitted Beyond the Limit: Introducing Mixedbread Wholembed v3 as a story · 2 points
HN discussionI'd love to know how they compare versus MixedBread, what relative strengths each has. https://www.mixedbread.com/ I really really enjoy & learn a lot from the mixedbread blog. And they find good stuff to open source (although the product itself is closed). https://www.mixedbread.com/blog I feel like there's a lot of overlap but also probably a lot of distinction too. Pretty new to this space of products though.
Comment linking /blog
on “ANN v3: 200ms p99 query latency over 100B vectors”Submitted Mixedbread: How We Built Multimodal Late-Interaction at Billion Scale as a story · 2 points
HN discussionSubmitted Fantastic (Small) Retrievers and How to Train Them as a story · 2 points
HN discussionSubmitted The Ceiling: How OCR Quality Limits RAG Performance as a story · 2 points
HN discussionSubmitted Baked-in Brilliance: Reranking Meets RL with mxbai-rerank-v2 as a story · 2 points
HN discussionSubmitted Building a Robust Ingestion System for Any File of Any Size as a story · 1 points
HN discussionSubmitted Toast 1 as a story · 1 points
HN discussionSubmitted Building a Robust Ingestion System for Any File of Any Size as a story · 1 points
HN discussionSubmitted Asymmetric Quantization: Near-Lossless Retrieval with 97% Storage Reduction as a story · 1 points
HN discussionSubmitted Dense Retrievers Know More Than They Can Express as a story · 1 points
HN discussionSubmitted A Delicious Free Lunch: Better Projections Improve ColBERT as a story · 1 points
HN discussion
