At the Deploy London 2025 event, we unveiled the exciting advancements in the Gradient AI Platform, marking a significant step forward in the realm of artificial intelligence (AI) for developers and enterprises alike. Our goal is to simplify the creation of AI applications that are ready for production, whether you’re just starting with your first AI agent or expanding an enterprise-level workload.
In line with this vision, we are thrilled to announce a suite of new features designed to broaden the possibilities of what users can create with the Gradient AI Platform. These enhancements not only offer improved security and control but also accelerate the development process, making it more efficient than ever.
AI applications are evolving to become more multimodal and data-centric, capable of processing text, images, audio, and a variety of other formats. To support this shift, we’ve introduced Image Model Support and Knowledge Base Auto-Indexing to the Gradient AI Platform. These features enable developers to provide their AI agents with a diverse array of inputs and information sources.
Image Model Support
One of the standout features is the new Image Model Support, which allows for the programmatic generation of images using text prompts. This is achieved through the Gradient AI Platform’s Serverless Inference API, which is powered by the gpt-image-1 model from OpenAI. This marks the platform’s first foray into non-text modalities, transitioning from text-only capabilities to include text-to-image generation.
Key aspects of this feature include:
- Text-to-image generation: Developers can create images directly via the Serverless Inference API.
- API integration: Image generation can be accessed using the existing inference endpoint.
- Unified billing: Charges for image generation appear on your DigitalOcean account alongside other services, simplifying financial management.
- Developer-first workflow: Enjoy direct API access without the need for UI-based tools or self-hosting.
- Enterprise authentication: Secure access is ensured through Model Access Keys.
This capability empowers users to generate images using natural language prompts, useful for a range of projects such as content creation, marketing materials, and product imagery.
For those eager to begin, detailed documentation is available to help you get started with Image Model Support.
Knowledge Base Auto-Indexing
Keeping AI agents current without manual effort is now easier with Knowledge Base Auto-Indexing. This feature automatically detects, retrieves, and re-indexes new and updated documents from connected sources into your OpenSearch database, ensuring your agent’s knowledge remains up to date.
The main functionalities include:
- Source connection: Easily integrate with platforms like Google Drive, Amazon S3, or Dropbox.
- Auto-indexing toggle: Activate auto-indexing and establish a default schedule.
- Customizable frequency: Select daily, weekly, or manual indexing options.
- Ingestion logic: Automatically manage new, updated, or deleted documents.
- Sync logs: Monitor the last sync time, status, and any errors for each source.
This automation streamlines the process of maintaining knowledge bases, saving time and reducing manual maintenance needs.
VPC Integration
As AI adoption expands, so do the requirements for privacy, compliance, and security. To address these needs, the Gradient AI Platform now includes Virtual Private Cloud (VPC) integration. This feature enables the establishment of secure, private network connections between services and resources within your own VPC, minimizing exposure to the public internet while maintaining deployments within DigitalOcean’s managed infrastructure.
Features of VPC integration include:
- Agent VPC connectivity: Connect agents within the platform to private customer databases.
- Egress IP visibility: Identify outbound IP addresses to configure your firewall.
- Cross-VPC support: Facilitate connectivity to multiple VPCs as required.
With this integration, you can securely run agents and indexing jobs with private network access, meeting security requirements without the burden of managing infrastructure.
Gradient AI Agent Development Kit (Private Preview)
The Gradient AI Agent Development Kit (ADK) is a code-first toolkit designed to facilitate the building, testing, and deployment of AI agent workflows directly from your development environment. This toolkit offers a framework for defining agent workflows, integrating models, and connecting tools or APIs, with insights and traces available within your workspace.
Notable features include:
- Expressive APIs: Manage deployments and define agent workflows entirely through code.
- Model integration: Connect both open-source and commercial models to your agents.
- Custom tools and connectors: Wrap APIs or functions for agent use.
- Traces and insights: Access agent session and workflow traces via your workspace interface or through code.
The Gradient AI Agent Development Kit will soon be available, offering early access to those who sign up for the private preview.
Gradient AI Genie (Private Preview)
Launching soon, Gradient AI Genie is a new experience integrated within Visual Studio Code. It enables users to interact with their agents using natural language directly from VS Code, aiding in the design, configuration, and connection of multi-agent systems.
Key functionalities include:
- Get started: Configure agents in VS Code with the Gradient AI ADK.
- Get help and recommendations: Receive workflow guidance, on-demand tips, and improved output.
- Automate: Orchestrate agents, build guardrails, and generate synthetic datasets.
Gradient AI Genie allows for the seamless building and orchestration of multi-agent systems without leaving the IDE. Sign up for the private preview to gain early access to this innovative feature.
In conclusion, these new features are poised to streamline the way users build, manage, and secure AI agents. Whether you’re generating images, maintaining up-to-date knowledge bases, or connecting workloads via private networks, the Gradient AI Platform offers robust tools to tackle increasingly complex AI workflows with ease. Additionally, the upcoming Gradient AI Agent Development Kit and Gradient AI Genie promise even more powerful tools for developing and orchestrating multi-agent systems directly from your development environment.
To explore these new features in depth, join our upcoming webinar, where we’ll provide a comprehensive walkthrough of these updates and share practical tips for building with the Gradient AI Platform.
For more Information, Refer to this article.
