Sold by: Cohere
Deployed on AWS
Embed 5 is Cohere's most powerful embedding family yet, delivering frontier retrieval on complex enterprise data that matches or exceeds the competition and makes major gains on Embed 4.
Overview
Every AI application is only as good as the information it retrieves. Embed 5 is Cohere's most powerful embedding family yet, turning reports, scanned pages, charts, contracts, and code into searchable representations of meaning, so the right content surfaces even when the words don't match. That means fewer more accurate answrs, fewer wasted LLM tokens, and better end user experiences.
Embed 5 Fast is for workloads where response time matters. Despite its compact size, it rivals or beats much larger models in finance, complex-document, and multimodal retrieval. Screen payments for fraud as they're authorized, refresh product results with every keystroke, route tickets instantly, or guide support agents during live calls.
Key specifications:
- 128K-token context window for long documents
- Embeds text, image, and mixed text-and-image pages
- Supports 100+ languages, including cross-language search
- Adjustable vector size dimensions (Matryoshka Representation Learning) with float, int8, and binary outputs
- Shared embedding space with Embed 5 Pro
- Works with OpenSearch, LangChain, Pinecone and more
Highlights
- Where pace meets performance. Embed 5 Fast is the highest-performing embedding model in its size class, built for interactive search, agent loops, and high-volume query traffic. Like Pro, Embed 5 Fast excels with professional, multimodal documents and regularly outperforms competitor flagship models at a fraction of the cost.
- Breakthrough embeddings speed.* Embed 5 Fast is optimized for low-latency, high-throughput retrieval. In our benchmarks, it delivered up to 71% lower p50 latency, 47% lower p99 latency, and 3.1x higher document throughput than Embed 5 Pro.
- Built to scale with Embed 5 Pro. Embed 5 Fast shares an embedding space with Embed 5 Pro, so both models work against the same vector index. That means one index to store, secure, and maintain, with no costly re-embedding as needs change. Route latency-sensitive traffic to Fast and quality-critical queries to Pro, or shift workloads between them over time. Balance performance, speed, and cost on one consistent retrieval setup.
Details
Sold by
Delivery method
Latest version
Deployed on AWS
New
Introducing multi-product solutions
You can now purchase comprehensive solutions tailored to use cases and industries.
Features and programs
Financing for AWS Marketplace purchases
AWS Marketplace now accepts line of credit payments through the PNC Vendor Finance program. This program is available to select AWS customers in the US, excluding NV, NC, ND, TN, & VT.
Pricing
Pricing is based on actual usage, with charges varying according to how much you consume. Subscriptions have no end date and may be canceled any time.
Additional AWS infrastructure costs may apply. Use the AWS Pricing Calculator to estimate your infrastructure costs.
Dimension | Description | Cost/host/hour |
|---|---|---|
ml.g4dn.12xlarge Inference (Batch) Recommended | Model inference on the ml.g4dn.12xlarge instance type, batch mode | $3.36 |
ml.p5.4xlarge Inference (Real-Time) Recommended | Model inference on the ml.p5.4xlarge instance type, real-time mode | $3.36 |
ml.g5.xlarge Inference (Real-Time) | Model inference on the ml.g5.xlarge instance type, real-time mode | $2.39 |
ml.g5.2xlarge Inference (Real-Time) | Model inference on the ml.g5.2xlarge instance type, real-time mode | $3.36 |
ml.g6.xlarge Inference (Real-Time) | Model inference on the ml.g6.xlarge instance type, real-time mode | $3.36 |
ml.g6.2xlarge Inference (Real-Time) | Model inference on the ml.g6.2xlarge instance type, real-time mode | $3.36 |
ml.g6e.xlarge Inference (Real-Time) | Model inference on the ml.g6e.xlarge instance type, real-time mode | $3.36 |
ml.g6e.2xlarge Inference (Real-Time) | Model inference on the ml.g6e.2xlarge instance type, real-time mode | $3.36 |
ml.g7e.2xlarge Inference (Real-Time) | Model inference on the ml.g7e.2xlarge instance type, real-time mode | $3.36 |
ml.g7e.4xlarge Inference (Real-Time) | Model inference on the ml.g7e.4xlarge instance type, real-time mode | $3.36 |
Dimensions summary
You pay by the hour for the SageMaker instance type you run this embedding model on. Pricing follows the compute capacity you choose, not the volume of text you process. One instance type, ml.g4dn.12xlarge, runs in batch mode for processing large sets of data at once. The remaining fifteen instance types run in real-time mode for immediate, on-demand responses. Within each mode, larger instance families carry higher hourly rates. You match the instance size to your workload and scale cost by selecting more or fewer host hours.
Top-of-mind questions for buyers
One host hour is one hour that a single SageMaker instance of the chosen type runs this model. Charges accrue per running hour, regardless of how much text you process. If you run several instances at once, each instance accrues its own host-hour charges in parallel.
Batch mode runs on the ml.g4dn.12xlarge instance and processes large data sets together. Real-time mode runs on the other fifteen instance types for immediate, on-demand responses. Both bill by host hours for the running instance. You pick the mode and instance type that fit your workload.
Software charges apply per host hour while the instance runs. A fully stopped or deleted endpoint stops accruing host-hour charges. Underlying AWS infrastructure or storage fees may still apply separately. The model license meters running time only, so idle but running instances still accrue charges.
cohere.com
Helpful?
Vendor refund policy
No refunds available.
Legal
Vendor terms and conditions
Upon subscribing to this product, you must acknowledge and agree to the terms and conditions outlined in the vendor's End User License Agreement (EULA) .
Content disclaimer
Vendors are responsible for their product descriptions and other product content. AWS does not warrant that vendors' product descriptions or other product content are accurate, complete, reliable, current, or error-free.
Delivery details
Amazon SageMaker model
An Amazon SageMaker model package is a pre-trained machine learning model ready to use without additional training. Use the model package to create a model on Amazon SageMaker for real-time inference or batch processing. Amazon SageMaker is a fully managed platform for building, training, and deploying machine learning models at scale.
Deploy the model on Amazon SageMaker AI using the following options:
Version release notes
Embed 5 is a state-of-the-art embedding model from Cohere with leading performance across financial, visually rich, and multimodal content. Embed 5 is built with enterprise documents and workloads in mind, enabling state-of-the-art retrieval performance across a variety of enterprise industries.
Additional details
Inputs
- Summary
This model accepts JSON requests that specifies a content object which can contain a list of texts, a list of data urls of a base64 encoded images or combination. This model supports interleaved images and texts in the same request. { ""content"": [ { ""type"": ""text"", ""text"": ""Look at my awesome car!"" }, { ""type"": ""image"", ""image"": f""data:image/png;base64,{base64_image}"" }, { ""type"": ""text"", ""text"": ""Do you want to buy it?"" }, ]
- Input MIME type
- application/json
{ ""content"": [ { ""type"": ""text"", ""text"": ""Look at my awesome car!"" }, { ""type"": ""image_url"", ""image_url"": ""data:image/png;base64,{base64_image}"" }, { ""type"": ""text"", ""text"": ""Do you want to buy it?"" } ], ""input_type"": ""search_document"", “embedding_types”: [""int8”], ""max_tokens"":8000, “output_dimensions”: 1024 }
Cohere's embedding models do NOT support batch transform.
Input data descriptions
The following table describes supported input data fields for real-time inference and batch transform.
Field name | Description | Constraints | Required |
|---|---|---|---|
inputs | A list of dicts with the key "content", which is a list whose type is either text or image_url. If type='text', there is a text key with a string. If type=image_url, it's a data URL formatted base64-encoded image and an image key. The max pixels per image is 2,458,624, the max memory size of a single request is 20 MB, and inputs objects can have at most 96 inputs. The inputs object can have a maximum of 800,000 tokens, calculated as follows: 1) For each image: image pixels / 784 (pixels per token) = tokens 2) For each text: text tokens 3) Total inputs object = for each input, the sum of (1) and (2). | The input data type is categorical. The default value is none. | No |
texts | An array of strings for the model to embed. Maximum number of texts per call is 96. If you are using the texts parameter, you cannot use the images parameter in the same call. The input data type is text. | N/A | No |
images | An array of base64-encoded data URLs as strings to embed.
There is no limit on the number of images per call. | The combined size of all images in the request must be at most 20MB. | No |
input_type | Specifies the type of input passed to the model.
"search_document": Used for embeddings stored in a vector database for search use-cases.
"search_query": Used for embeddings of search queries run against a vector DB to find relevant documents.
"classification": Used for embeddings passed through a text classifier.
"clustering": Used for the embeddings run through a clustering algorithm.
"image": Used for embeddings with image input. | Value must be one of: search_document, search_query , classification, clustering, or image | Yes |
embeddings_type | Specifies the types of embeddings you want to get back. Not required. If unspecified, defaults to float . | Input must be one of: float , int8 , uint8, binary, ubinary, base64 | No |
output_dimension | Specifies the length of the output dimensions of the embeddings vector. Not required. If unspecified, returns 1024 dimensions per embedding vector. | Input data type is categorical. If categorical is chosen: {256, 512, 768, 1024, 1536, 2048}. The default value is 1024. | No |
max_tokens | The maximum number of tokens considered for each input object before it is truncated; the default for this model is set at 128,000 tokens. | The input data type is integer. | No |
truncate | One of NONE, START, or END to specify how the API will handle inputs longer than the maximum token length. Passing START will discard the start of the input. END will discard the end of the input. In both cases, input is discarded until the remaining input is exactly the maximum input token length for the model. If NONE is selected, an error will be returned when the input exceeds the maximum input token length. | The input data type is categorical. The default value is NONE. | No |
Resources
Vendor resources
Support
Vendor support
AWS infrastructure support
AWS Support is a one-on-one, fast-response support channel that is staffed 24x7x365 with experienced and technical support engineers. The service helps customers of all sizes and technical abilities to successfully utilize the products and features provided by Amazon Web Services.
Similar products
Embed 5 is Cohere's most powerful embedding family yet, delivering frontier retrieval on complex enterprise data that matches or exceeds the competition and makes major gains on Embed 4.
Cohere provides a multilingual representative AI model, that translates texts and images into numerical vectors that models can understand.
Cohere provides a representative AI model: Embed that translates text into numerical vectors that models can understand.
Cohere provides a representative AI model, Embed Light, that translates texts and images into numerical vectors that models can understand.
Customer reviews
No customer reviews yet
Be the first to review this product . We've partnered with PeerSpot to gather customer feedback. You can share your experience by writing or recording a review, or scheduling a call with a PeerSpot analyst.