The best part for me is the out-of-the-box Retrieval-Augmented Generation (RAG) and API integration capabilities. It lets developers ground AI responses in trusted enterprise data and dynamically call external tools, without having to build complex custom middleware from scratch. Review collected by and hosted on G2.com.

Customizing underlying model behaviors can feel restrictive at times, and troubleshooting multi-step agent workflows often lacks transparency. It’s not always clear what’s happening at each step, which makes it harder to diagnose issues or fine-tune the overall flow. Review collected by and hosted on G2.com.

very simple and easy to connect with, hence very useful for day to day work Review collected by and hosted on G2.com.

Well, I think the pricing should be a little cheaper, and the context awareness still needs to improve. Review collected by and hosted on G2.com.

It's not easy to use, but is very versatile to integrate with your project. Also, you can choose between modules and get some new ideas about it. The interface is very, very good in terms of testing the module right away. Review collected by and hosted on G2.com.

It's just not as easy as the hugging face is smolagents. Review collected by and hosted on G2.com.

I used Vertex AI for image recognition and generation, and I really like that it provides everything in one place for data preparation, training models, and deployment. It's the best tool for ML projects, and it works smoothly. We use it mainly for deployment and training models, and I love that aspect as it makes our work fast and easy. As Google developers, our work is in Google Cloud, so Vertex AI is a major tool for us to train models. Its updated GPT makes the work fast and easy, and we feel comfortable using it. It's also user-friendly, and even a beginner can learn the setup easily, which makes it feel unique. Review collected by and hosted on G2.com.

We use Vertex AI only in Google Cloud, and while it works fine there, when connecting to some other things, we feel that we don't get proper output. It would be better if there was more interaction with the AI to make it more user-friendly. Review collected by and hosted on G2.com.

I like that Vertex AI offers a playground to test out models and different inputs. It's an easy method to try out different prompts and instructions to the agents without having to use any code. This makes it easier for me to understand which models work better for my use case. Review collected by and hosted on G2.com.

I think that the traceability and monitoring of users can be improved. It would be better if I could add my own MCP servers and have it test connections and users of the tools themselves. Initially, setup wasn't easy because I had to use offline Google credentials stored in a JSON file, which made it difficult for my applications. But later on, using an API key made things easier. Review collected by and hosted on G2.com.

I like that Vertex AI brings the whole ML workflow into one platform and integrates well with Google Cloud services. It also saves time by handling infrastructures and scaling automatically. I also like how easy it is to deploy models and manage them through APIs. The platform is flexible and works well for both experimentation and production workloads. Review collected by and hosted on G2.com.

One area that could be improved is the learning curve for new users, especially when configuring services in Google Cloud. Pricing and documentation could also be clearer for beginners. Review collected by and hosted on G2.com.

I like that Vertex AI allows me to build, train, and deploy machine learning models without dealing with heavy infrastructure. It efficiently handles infrastructure and environment setup, saving me from the hassle of managing servers or configurations. I appreciate that all things are in one place, including data prep, model training, experimentation, deployment, and monitoring, which are all connected. Experimentation feels smoother because I can explore data, test models, and iterate quickly in a managed notebook environment without worrying about setup or performance issues. It's reliable and saves a lot of time when moving back and forth between analysis and modeling. Review collected by and hosted on G2.com.

There are a lot of features packed in, which is powerful, but it also means simple tasks sometimes take longer to find or configure than expected. A more streamlined experience for common workflows would help. Review collected by and hosted on G2.com.

I like Vertex AI's unified and production-ready approach to machine learning. It brings model training, deployment, monitoring, and access to foundation models into a single platform, significantly reducing operational overhead. The scalability, tight integration with Google Cloud services, and support for both custom models and generative AI make transitioning from experimentation to real-world deployment easy. I also appreciate how it solves the challenge of managing the end-to-end machine learning lifecycle in production, reducing the complexity of model training, deployment, scaling, and monitoring. Its unified platform also addresses infrastructure management, experiment tracking, and collaboration between data science and engineering teams, enabling faster and more reliable delivery of AI solutions. Plus, the initial setup was pretty smooth and easy. Review collected by and hosted on G2.com.

While Vertex AI is powerful, it can feel complex for new users due to the number of services and configuration options involved. Costs can also be difficult to predict, especially when running experiments or scaling models. I’d like to see improvements in onboarding and usability, especially clearer guidance for transitioning from experimentation to production. More intuitive cost visibility and real-time spend alerts would help teams manage budgets more effectively. Review collected by and hosted on G2.com.

I use Vertex AI as an all-in-one place to handle our machine learning work from start to finish, which lets us prepare data, build and train models, test different ideas, and put models into production in a more organized way. It makes teamwork smoother as data scientists and engineers can work from the same workflows without a lot of back-and-forth. Vertex AI solves the problem of fragmented and time-consuming machine learning workflows by bringing these steps together in one place, reducing manual effort and confusion. It keeps experiments, versions, and results organized, which makes collaboration easier. I appreciate how it simplifies monitoring and managing models once they’re in production with built-in tools for tracking model performance, detecting drift, and analyzing predictions. The ability to run multiple experiments in parallel is extremely helpful, and the platform’s collaborative features allow team members to work together, share insights, and maintain consistency. Vertex AI feels reliable and well-structured, helping the team work faster and with more confidence. Review collected by and hosted on G2.com.

While I really value what Vertex AI offers, there are a few areas where it could be smoother. Sometimes the interface and navigation feel a bit overwhelming, especially when juggling multiple complex projects or if you’re just getting started. Some of the advanced features, like customizing pipelines or tweaking deployment settings, have a steep learning curve and require digging through documentation, which can slow things down. Certain resource-heavy workflows can also feel a bit slow unless you optimize carefully. The reporting and visualization tools are useful, but I wish they were a bit more intuitive and flexible so I could get insights faster without relying on external tools. Overall, making navigation simpler, guiding users more through advanced features, and improving performance and reporting flexibility would make Vertex AI even easier and more efficient to use. Review collected by and hosted on G2.com.

I really like Vertex AI for its JupyterLab environment, which is incredibly useful for development. Terminal access and the real-time file manager provide great flexibility, especially with their upload and download options. The cloud platform is convenient for connecting and using various resources needed for machine learning and deep learning model development. I find the different RAM, memory, GPU, and CPU compute options essential for my tasks. I appreciate the built-in editors for JSON and xlsx files because I don't need to manually install different extensions for viewing and editing these formats. Overall, the simplicity of the initial setup stood out to me. Review collected by and hosted on G2.com.

The UI can be improved. It does not support autocomplete or code suggestions. It lacks features available in VS Code IDE like multi-cursor support and rich code editing, which could enhance productivity. The UI only offers a white color option, and it could benefit from a variety of color options. Extensions for editing Vertex AI hosted codebases directly from VS Code would be helpful. Review collected by and hosted on G2.com.