We are finishing the year with a ton of new products, features and improvements with Clarifai Release 7.0.

## New API Clients for Swift, Rust, Go and C++

There are more ways than ever to use the [Clarifai](/content/site-root.html) API with our new Swift, Rust, Go and C++ clients. Our clients are auto-generated using [gRPC](/content/blog/muxing-together-grpc-and-http-traffic-with-grpc-gateway/index.html), so they will always have the latest available [Clarifai API](/content/blog/clarifai-api-conventions/index.html) feature-set. [Get started using these new clients today.](https://docs.clarifai.com/api-guide/api-overview/api-clients)

## Bulk Editing of Concepts, Metadata and Geodata

Clarifai is introducing new and improved ways of labeling your data right in Data Mode. You can add concepts, [metadata](/content/blog/how-visual-similarity-and-custom-metadata-can-enhance-your-search/index.html) and geodata to your inputs at upload time, or select multiple inputs in Explorer and label your data there. [Learn more about labeling your data in bulk.](https://docs.clarifai.com/legacy-portal-guide/data/explorer/annotations)

## AI Assist with Public Models

Supercharge your labeling [workflows with AI Assist](/content/products/scribe-data-labeling-platform/index.html). It is faster and easier than ever to label your data with the power of [Clarifai models](/content/blog/model-mode/index.html). Use our knowledge graph to connect the concepts in our prebuilt models to the concepts in your custom models. [Find out more about using AI Assist for your labeling tasks.](https://docs.clarifai.com/portal-guide/annotate/ai-assist)

## Text Labeling in Scribe

Scribe now supports text labeling, so that you can label text and create your own [custom NLP models](/content/products/natural-language-processing/index.html), and build powerful multi-modal workflows. [Check it out.](https://docs.clarifai.com/portal-guide/annotate/labeling-tools)

## Manual Review - Annotator View

When manually reviewing labels, annotator view provides a grid layout that makes it easy to spot check, or individually review the labels provided by your labelers. Labels are grouped by concept for convenience. [Learn more about reviewing labeling tasks in our documentation.](https://docs.clarifai.com/portal-guide/annotate/review)

## EfficientDet DeepTrain

You can now DeepTrain models using our [cutting edge implementation](/content/blog/bring-digital-asset-management-dam-into-the-future-with-automatic-image-tagging-and-keywording/index.html) of the EfficientDet algorithm. EfficientDet provides a [neural network](/content/blog/neural-network-pruning-for-compression-understanding/index.html) architecture that significantly improves efficiency and accuracy for object detection. [Learn more about deep training.](https://docs.clarifai.com/portal-guide/model/deep-training)

## EfficientDet for People Detector V2

EfficientDet is the basis for our new People Detector V2 model and it runs 3x faster. [Try it yourself in Portal.](/content/products/ai-portal/index.html)

## Public Model Deprecations

Please note that we have deprecated the following public models:

| Model Name                       | Model ID                         | New Model or Workflow       |
|----------------------------------|----------------------------------|-----------------------------|
| demographics                     | c0c0ac362b03416da06ab3fa36fb58e3 | demographics workflow        |
| apparel-detect-embed             | 71a2ef27d4ff8a6b096394de03e7d2d5 |                             |
| apparel-detect-apparel-concept   | a8436308d0f92d2d5b0719816ee78793 |                             |
| OCR Detection                     | 9bff11543fd7e3d3fb215f30ef927ca3 | VTR workflow                |

Learn more about our releases in the [Changelog](https://docs.clarifai.com/product-updates/changelog).