Google Cloud enables self-deployment of proprietary AI models in customer VPCs
Google Cloud has announced that users can now self-deploy proprietary AI models from partners such as Mistral AI and CAMB.AI within their own Virtual Private Cloud (VPC) on Vertex AI. This update allows streaming and tech enterprises to run closed-source models for various tasks, including multilingual text-to-speech, while maintaining data sovereignty and VPC security policies. The models are available through the Vertex AI Model Garden, offering a curated catalog for discovery, testing, and deployment.
Key Takeaways
- Users can now self-deploy proprietary models from eight partners, including Mistral AI and CAMB.AI, directly into their VPC.
- This capability supports closed-source models and those with restricted commercial licenses, maintaining adherence to VPC-SC policies.
- The Vertex AI Model Garden serves as a curated catalog for discovering, testing, and deploying these models, which include AI21 Labs' Jamba Large 1.6 and CAMB.AI's MARS7 for hyper-realistic TTS.
- Organizations can optimize performance or cost by scaling deployments manually or with auto-scaling, and select Google Cloud regions for data compliance.
Why It Matters
This move by Google Cloud significantly addresses enterprise demand for greater control and security over AI model deployment, particularly for sensitive streaming data. Allowing proprietary models within a customer's VPC mitigates data residency and compliance concerns, which are critical for media companies handling vast amounts of user information and localized content. It enables faster adoption of advanced AI capabilities, like multilingual voice cloning, without compromising data security. The next indicator will be the rate at which major streaming platforms adopt this self-deployment feature for their AI workflows, signaling a shift in cloud AI strategy.
Read full article at cloud.google.com