Jina AI | Latest News & Updates - Nov 27, 2024 Release

Jina AI announced that their jina-embeddings-v3 model can create text watermarks that persist after paraphrasing and translation, as presented at EMNLP2024...


Brought to you by RivalSense - an AI tool for monitoring any company.

RivalSense tracks the most important product launches, fundraising news, partnerships, hiring activities, pricing changes, tech news, vendors, corporate filings, media mentions, and other developments of companies you're following 💡


Jina AI

🌎 jina.ai

Founded in 2020 in Berlin, Jina AI is a leading AI company specializing in search technology. The company offers the Search Foundation, which includes embeddings, rerankers, and other search-focused models, aimed at empowering businesses and developers to unlock the value of multimodal data through enhanced search capabilities.


Jina AI - Latest News and Updates

  • Jina AI announced that their jina-embeddings-v3 model can create text watermarks that persist after paraphrasing and translation, as presented at EMNLP2024.
  • Jina AI's CLIP V2 can now be deployed using LitServe as detailed in a guide available on Lightning AI Studio.
  • Jina AI has been integrated into ChatCraft for PDF to Markdown conversion, enhancing its document processing capabilities.
  • Jina AI released a vision encoder model with NC licensed weights, potentially for document image retrieval.
  • Jina AI is working on RAG evaluations to identify weaknesses in RAG pipelines using synthetic data.
  • Jina AI has developed a new corpus for pairwise contrastive training to refine embeddings in Elasticsearch.
  • Scott Martens is a Senior Content Manager and Evangelist at Jina AI in Berlin.
  • Jina AI's CEO Han Xiao explained that the jina-embeddings-v3 model performs poorly for visualization at low dimensions like 2 or 3 because it wasn't trained for such dimensions and isn't optimized for visualization-specific criteria.
  • Zhichao Xu Brutus presented two posters and gave a talk at EMNLP invited by Jina AI about using the Mamba model for ranking.

Sign up to receive regular updates


If you liked the insights, consider following your own companies of interest. Receive weekly insights directly to your inbox using RivalSense.