Cohere Launches Embed 4: Advanced Multimodal AI Model for Enterprise Search and Retrieval

Cohere launches Embed 4: New multimodal search model processes 200-page documents

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More Enterprise retrieval augmented generation (RAG) remains integral to the current agentic AI craze. Taking advantage of the continued interest in agents, Cohere released the latest version of its embeddings model with longer context windows and more multimodality. Cohere's Embed 4 builds on the multimodal updates of Embed 3 and adds more capabilities around unstructured data. Thanks to a 128,000 token context window, organizations can generate embeddings for documents with around 200 pages. "Existing embedding models fail to natively understand complex multimodal business materials,‬‭ leading companies to develop cumbersome data pre-processing pipelines that only slightly‬‭ improve accuracy," Cohere said in a blog post. "Embed 4 solves this problem, allowing enterprises and their employees to‬‭ efficiently surface insights that are hidden within mountains of unsearchable information.‬" Enterprises can deploy Embed 4 on virtual private clouds or on-premise technology stacks for added data security. Companies can generate embeddings to transform their documents or other data into numerical representations for RAG use cases. Agents can then reference these embeddings to answer prompts. Domain-specific knowledge Embed 4 "excels in regulated industries" like finance, healthcare and manufacturing, the company said. Cohere, which mainly focuses on enterprise AI use cases, said its models consider the security needs of regulated sectors and have a strong understanding of businesses. The company trained Embed 4 "to be robust against noisy real-world data" in that it remains accurate despite the "imperfections" of enterprise data, such as spelling mistakes and formatting issues. "It is also performant at searching over scanned documents and‬ handwriting. These formats are common in legal paperwork, insurance invoices, and expense‬ receipts. This capability eliminates the need for complex data preparations or pre-processing‭ pipelines, saving businesses time and operational costs," Cohere said. Organizations can use Embed 4 for investor presentations, due diligence files, clinical trial reports, repair guides and product documents. ‭ The model supports more than 100 languages, just like the previous version of the model. Agora, a customer of Cohere, used Embed 4 for its AI search engine and found that the model could surface relevant products. "E-commerce data is complex, containing images and multifaceted text descriptions. Being able to‬ represent our products in a unified embedding makes our search faster and our internal tooling more‬ efficient," said Param Jaggi, Founder of Agora‬, in the blog post. Agent use cases Cohere argues that models like Embed 4 would improve agentic use cases and claims it can be "the optimal search engine" for agents and AI assistants across an enterprise. "In addition to‬‭ strong accuracy across data types, the model delivers enterprise-grade efficiency," Cohere said. "This allows it‬ to scale to meet the demands of large organizations." Cohere added that Embed 4 creates compressed data embeddings to cut high storage costs. Embeddings and RAG-based searches let the agent reference specific documents to fulfill request-related tasks. Many believe these provide more accurate results, ensuring the agents do not respond with incorrect or hallucinated answers.

Cohere releases Embed 4: a multimodal AI model designed for agentic search - SiliconANGLE

Cohere releases Embed 4: a multimodal AI model designed for agentic search Artificial intelligence startup Cohere Inc. today launched Embed 4, its latest AI model designed to provide embeddings for search and retrieval for AI applications such as assistants and agents. Enterprise businesses rely heavily on retrieval augmented generation, a technique that allows generative AI models to retrieve and incorporate fresh, accurate information in real-time so that large language models respond to user queries with the best possible data. Embedding models, such as Embed take data from documents and turn it into mathematical representations called vectors that can be used to represent the dynamic, multifaceted context of the information. In the case of Embed 4, this can include anything from text-based documents or images to tables, graphs, code and diagrams. Cohere said the new model includes an extremely large context length - up to 128,000 tokens, or around a 200-page document - allowing it to ingest a lengthy annual financial report, product manual or detailed legal contract. It is also deeply multilingual with over 100 languages, featuring key business languages including Arabic, Japanese, Korean and French proficiencies in addition to English. "Embed 4 enables organizations to search their unstructured documents, where a large majority of their important data resides," Cohere said about Embed 4 in the announcement. "It is uniquely capable of generating high-quality representations of complex mixed-modality documents - all within a unified vector." The AI startup said it designed the new model to excel in regulated industries such as finance, healthcare and manufacturing with domain-specific understanding of these industries. These include capabilities for searching investor presentations, annual financial reports, medical records, procedural charts, product specification documents, repair guides and supply chain documents. Cohere also noted that Embed 4 can deal with noisy real-world data by handling fuzzy images and poorly oriented documents. The model was trained against numerous documents of scanned documents, handwriting and other distressed documents, the company said. These are the types of complex data that many businesses will encounter in day-to-day multimodal data pre-processing which are part of the human manual pipeline. Agora, an AI-driven search engine for 35,000 online stores and a customer of Cohere used the model to assist with its business, said it was able to build a better search using its advanced multimodal embedding features. "E-commerce data is complex, containing images and multifaceted text descriptions," said Param Jaggi, founder of Agora. "Being able to represent our products in a unified embedding makes our search faster and our internal tooling more efficient," Embed's capabilities are essential for accurate search and retrieval, which power generative AI models such as Cohere's Command A, a low-cost model the company released last month. Models such as Command A power conversational assistants and AI agents, but rely heavily on search engines, which are connected to secure, proprietary company information to source relevant information to user questions. This is necessary to speed up responses, increase accuracy and reduce hallucinations. Cohere said the new Embed 4 model is integrated with North, the company's secure AI agent productivity platform where it powers its semantic search capability in its Compass product. The Embed 4 model is also available starting today from Microsoft Azure AI Foundry, on Amazon SageMaker and for private deployments.

Cohere Launches Embed 4 for Advanced Multimodal Enterprise Search | AIM Media House

Security and deployment flexibility are central to Embed 4's design. Enterprise AI startup Cohere has released Embed 4, a powerful new embedding model designed to help organisations search and make sense of large volumes of unstructured data, including documents combining text with visuals. Tailored for enterprise use, Embed 4 supports applications like AI assistants and agents that need a deeper understanding of a company's internal knowledge. It can process documents up to 1,28,000 tokens in length, roughly 200 pages, making it ideal for handling long-form content such as annual reports, legal agreements or technical documentation. The model supports more than 100 languages, including widely used business languages like Arabic, French, Japanese and Korean. This makes it easier for global teams to find and access the information they need, regardless of language. Security and deployment flexibility are central to Embed 4's design. It can be run within virtual private clouds or on premises, helping organisations regulate sectors such as finance, healthcare, or manufacturing meet strict data privacy and compliance requirements. One of Embed 4's standout strengths is its ability to deal with real-world document imperfections. Whether formatting issues, typos or scanned images are involved, the model is built to handle messy data often found in business documents such as invoices, charts or handwritten notes. It also excels at representing multimodal content -- documents that mix text with images, tables, graphs and even code -- in a single embedding. This improves the accuracy of enterprise search tools and cuts down on the need for heavy pre-processing when analysing complex files. Embed 4 is a key component in retrieval-augmented generation (RAG) pipelines, where generative models like Cohere's Command R rely on high-quality data retrieval before producing responses. As the engine behind this retrieval, Embed 4 helps ensure that generated content is more reliable, grounded in the right context and less prone to hallucination. Embed 4, along with Cohere's Command A language model, is now available via the Azure AI Foundry model catalogue.

Cohere's Embed 4 Can Search and Retrieve Internal Data for Businesses

Cohere released Embed 4, an artificial intelligence (AI) embedding tool, last week. The Toronto-based AI firm pitched the new search and retrieval tool to businesses building and deploying AI applications and agents. The company, which builds enterprise-focused AI models and tools, claims that Embed 4 can understand complex, multimodal documents and efficiently surface information that an AI system requires to complete a task. The AI model is also said to help businesses save on data storage costs by sharing compressed embeddings instead of full documents. In a blog post, Cohere announced the launch of Embed 4 and detailed the new product. It is a multimodal embedding tool that adds search and retrieval capability to existing AI systems. The tool is currently available directly from the company's website, Microsoft Azure AI Foundry, and Amazon SageMaker. It is also available for private deployment into any virtual private cloud (VPC) or on-premise environment. All AI models use a system dubbed Retrieval-Augmented Generation (RAG) to find information from their knowledge base. Essentially, it is a command that prompts search and retrieval of particular information based on keywords, ranking, and other rule-based algorithms. Embed 4 is essentially an AI model that replaces this function for data from outside sources. Cohere says the embedding tool can be added to any existing AI system, be it an AI application or an agent. Enterprises that usually use such tools internally, either use the third-party AI model's search engine or custom build search engines. The AI firm claims that Embed 4 is a better option than either of those two solutions. Embed 4's biggest unique proposition is its multimodality support. It can contextually understand documents that not only contain text, but also those that contain images, graphs, tables, diagrams, and code. Additionally, the AI tool supports more than 100 languages, including Arabic, Japanese, Korean, and French, to let businesses globally seamlessly look up their data. Cohere also highlighted that Embed 4 was trained against noisy real-world data, which means imperfect documents, including those with spelling mistakes, formatting issues, or different page orientation and can also be retrieved by the AI tool without harming the accuracy of the search results. Additionally, the AI model is equipped with domain-specific understanding of data from regulated industries such as finance, healthcare, and manufacturing. This means Embed 4 can be deployed in VPC and on-premise environments to keep data secure.