Mixedbread

Introducing Mixedbread Wholembed v3, our new SOTA retrieval model across all modalities and 100+ languages. Wholembed v3 brings best-in-class search to text, audio, images, PDFs, videos... You can now get the best retrieval performance on your data, no matter its format.

Read more here: mixedbread.com/blog/listwise-…

Introducing mxbai-rerank-v3-listwise: reranking that goes beyond binary relevance. It reads the whole candidate set, resolves conflicts, and ranks by directives like recency, source priority, and multi-step rules. +11% NDCG@10 on average across multiple domains, modalities, and languages in runs with Wholembed v3. Available today in preview in Mixedbread.

Mixedbread search's ultimate aim is to power all workflows, no matter their modality or language. Try it for your own knowledge-intensive tasks today: mixedbread.com

You can read more about this in our blog post, where we present more detailed benchmark results and elaborate on the nature of the three benchmarks, and why we're very proud to be topping all three of them. mixedbread.com/blog/closing-g…

So what is the Oracle gap? Optimising agentic systems is complicated. There are many individual components you need to get just right. Retrieval is one of those components, and its impact is best measured by the Oracle gap: the difference between the performance of the same system between an imperfect retriever and perfect, fully-relevant results that would be provided by a so-called Oracle.

Agents are increasingly performing knowledge work: Deep Research, generating financial reports, reasoning across historical knowledgebases... Many high-quality benchmarks now focus on evaluating such tasks, among which BrowseComp-Plus, @databricks's OfficeQA, or @Snowflake's MADQA, released just last week.

For Agentic tasks, Oracle-level performance is the maximum performance a system can achieve, assuming it is able to retrieve all relevant documents perfectly, every time. We're proud to show that Mixedbread Search approaches the Oracle on multiple knowledge intensive benchmarks.

I've been eagerly awaiting this release from the @mixedbreadai folks. They're world-leading experts in late interaction retrieval. And today they remind us that late interaction done well makes all your favorite embedding models look like they don't work.

Find out more about the model and its performance here: mixedbread.com/blog/wholembed…

Wholembed v3 is available immediately through Mixedbread Search. You can try it on our platform now, for free: New users get 2M free tokens to get started. Startups can receive much more through our partnered accelerator programs with Vercel and TinyFish. mixedbread.com

Introducing Mixedbread Wholembed v3, our new SOTA retrieval model across all modalities and 100+ languages. Wholembed v3 brings best-in-class search to text, audio, images, PDFs, videos... You can now get the best retrieval performance on your data, no matter its format.

The SIGIR deadline has now passed. You know what hasn't passed? The Late Interaction Workshop deadline! Which we are extending to February 20th AOE. We care about everything around multi-vector retrieval: indexing, field reports, model training, multimodal, you name it.

We're building Mixedbread to close the gap between Search that is possible today and what the users of tomorrow will demand. You can read more about it here: mixedbread.com/blog/multimoda…

Retrieval has always been, and continues to be the natural interface to information. Traditionally, it was how you found useful websites and relevant snippets. Nowadays, it's how agents find exactly the pieces of context they need to answer a user's query. But retrieval has been stagnant for too long, and it has become clear that the once-omnipresent single-vector text representation is not meeting the needs of this new generation of users. Agents need the model to be able to understand long, reasoning-intensive queries. They need the ability to retrieve documents where they live, be they text, images, pdfs or even videos.

We build the first production ready multi-vector and multimodal search. Now we are serving over 1 billion documents in under 50ms latency (p50). We are sharing how we build it.