SAP AI Core Reviews & Product Details

What I like most about SAP AI Core is how it supports AI workloads in a structured, scalable way. It offers a reliable environment for developing, deploying, and managing AI solutions, and it integrates smoothly with enterprise data and applications. I also appreciate the flexibility it provides across different AI models and workflows, which makes it practical for building and operationalizing a range of AI use cases. Review collected by and hosted on G2.com.

One area that could be improved is the learning curve, particularly for users who are new to SAP AI Core and its broader ecosystem. Setting up and configuring AI workloads can sometimes feel complicated, and it may take extra time to understand how the different components fit together. Clearer documentation, a simpler onboarding process, and more direct troubleshooting guidance would make the overall experience much easier for new users. Review collected by and hosted on G2.com.

I use SAP AI Core to run machine learning workflows and deploy models, including support request classification. I appreciate its support for managing the AI lifecycle within SAP BTP and for working with open-source machine learning frameworks on a platform designed to operate AI workloads. It’s useful for projects already connected to the SAP ecosystem. Having a consistent way to run workflows and manage deployments helps organize the move from building a model to making it available to a business application, and gives the project a clear structure. I find the SAP AI Launchpad helpful for our team. Moving from local experimentation to a managed deployment has made running workflows easier and more trustworthy, and the configuration has been easier and more reliable for applications to authenticate, send requests, and handle responses or failures. A working endpoint is valuable to me. The initial setup was easier than I expected: connecting to SAP BTP, using the documentation and Launchpad connection process, and adding the SAP AI Core service key to our existing application was enough to get started. Review collected by and hosted on G2.com.

I found the learning curve challenging as someone new to SAP BTP. I’d appreciate a clearer beginner journey from configuring account permissions to the first successful deployment, along with more context for setup errors, especially when several configuration choices could explain why something isn’t working. Review collected by and hosted on G2.com.

I like that SAP AI Core makes managing and deploying AI models easier. I like having a centralized platform to handle AI workflows without managing every part manually. I can manage AI workflows in one place, track deployments more easily, save time on repetitive tasks, and keep things organized across different AI projects. It works well with the SAP ecosystem, which makes integrating AI into existing business processes more convenient. It also fits into our existing workflows and helps us connect AI services with other systems. Review collected by and hosted on G2.com.

I found the initial setup and configuration fairly straightforward, but it took some time to understand and get everything working properly. I’d like more practical examples in the documentation to help new users get started quickly, along with better error messages and easier troubleshooting. Review collected by and hosted on G2.com.

What I like most about SAP AI Core is that it makes it easier to build, deploy, and manage AI models in a structured way. I especially like the integration with SAP’s ecosystem and the ability to manage AI workloads without having to handle all the infrastructure manually. Review collected by and hosted on G2.com.

The main thing I dislike about SAP AI Core is that it can feel a bit complex to set up and get started with, especially if you’re new to the SAP ecosystem. The documentation and configuration can also be overwhelming at times, and some workflows could be more straightforward. Review collected by and hosted on G2.com.

What I like most about SAP AI Core is how naturally it fits into the SAP BTP ecosystem. It offers a managed, Kubernetes-based runtime for training and serving AI models, which means we don’t have to run or maintain our own infrastructure. The Generative AI Hub is particularly valuable because it provides access to multiple large language models from different providers through a single, consistent interface, while also supporting enterprise-grade security and governance.

I also appreciate that configuration is Git-based, keeping workflows and serving setups version-controlled and reproducible. Resource groups make it easy to separate projects and tenants cleanly. SAP AI Launchpad gives the team a clear UI for monitoring executions, deployments, and metrics, and the Python SDK plus REST APIs make automation straightforward. Most importantly, SAP AI Core helps us bring AI directly into our SAP business processes and data, making it much easier to move from experimentation to production in a compliant way. Review collected by and hosted on G2.com.

The biggest drawback for me is the steep learning curve. Getting SAP AI Core set up involves quite a few steps: configuring the BTP subaccount and service keys, connecting a Git repository and Docker registry, creating secrets, and then writing Argo workflow and KServe serving templates. For teams that are new to Kubernetes concepts, this can quickly feel overwhelming.

The documentation is thorough, but it’s spread across SAP Help, tutorials, and blog posts, so tracking down the right example for a specific scenario often takes longer than expected. On top of that, error messages during executions and deployments aren’t always very clear, which slows down debugging.

Pricing and resource plans can also be difficult to estimate upfront, and costs may end up being high for small teams or early-stage experiments. Lastly, new models and features sometimes arrive in the Generative AI Hub later than they do directly from the model providers, and availability can vary by region. Even so, SAP does seem to be steadily improving the tooling and the onboarding experience. Review collected by and hosted on G2.com.

The most helpful and best thing about SAP AI Core is that it's really helpful in managing and deploying AI and Machine Learning Models. Also, it provides integration with the SAP Ecosystem for business integration. The best of all is that it helps in monitoring and maintaining the AI Life Cycle. Review collected by and hosted on G2.com.

Initial setup and configuration require technical knowledge and can be more simplified to enrich the user experience. The overall performance monitoring is okay but can be more detailed, with clearer error messages and more detailed and user-friendly documentation. There are no pricing/ROI examples. Practical pricing examples and real-world ROI scenarios would help to better understand costs, evaluate the value of their AI workloads, and can be helpful in optimizing resources. Review collected by and hosted on G2.com.

I like that SAP AI Core provides a centralized, scalable platform for managing AI workloads. Its integration with the SAP ecosystem, SAP BTP, and existing SAP applications makes it easier to bring AI capabilities into business applications and processes. I like the flexibility to work with different models and generative AI services while maintaining enterprise-level security and governance. I can manage the AI lifecycle from experimentation and model deployment to inference and monitoring without building all the infrastructure myself. Centralized access to different models makes development easier, saves development time, and makes it easier to move AI use cases into production. I can build and deploy AI components in SAP AI Core and integrate the results through APIs and other BTP services, which reduces custom integration and makes security and access management more consistent. Review collected by and hosted on G2.com.

I find the learning curve and ease of setup could be improved. SAP AI Core is powerful, but the initial configuration and understanding of concepts like scenarios, configurations, deployments, and resource groups can take some time. The setup is manageable, but not completely straightforward because it involves BTP configuration, service instances, roles and permissions, resource groups, and connections to other services. I think SAP could provide more guided, step-by-step onboarding with ready-to-use templates and complete sample projects. Review collected by and hosted on G2.com.

SAP AI Core is essentially an enterprise AI execution and management platform. It helps organizations deploy and operate ML and generative-AI workloads at scale, provides access to multiple foundation models through the Generative AI Hub, and adds enterprise capabilities like governance, orchestration, monitoring, data masking and grounding. Its main advantage is that developers can build AI applications without having to manage the entire model-serving and infrastructure layer themselves Review collected by and hosted on G2.com.

The main downside of SAP AI Core is that it can introduce additional complexity and cost. For simple AI applications, directly using an LLM provider may be easier and cheaper. AI Core becomes more valuable when you need enterprise-scale deployment, governance, security, SAP integration, and management of multiple AI workloads Review collected by and hosted on G2.com.

What I love most is that it saves you from juggling half a dozen AI vendor contracts by bringing major LLMs together behind one clean API. It also plugs directly into your business data and security rules, without forcing you to hack together risky, duct-taped pipelines just to make things work. And because it runs on standard Kubernetes and can scale to zero, you still get solid enterprise-grade capability without burning cash on idle compute. Review collected by and hosted on G2.com.

It buries developers in BTP cockpit red tape, service keys, and high enterprise minimums before they can write a single line of code. Unless you’re already deeply reliant on S/4HANA data, the developer experience feels clunky, the tooling lags behind, and it just isn’t worth the overhead. Review collected by and hosted on G2.com.

Access to many foundation models from various vendors in one place. This is the main benefit I think. Beside this, the Grounding feature is easy, and valuable. Access to GPU helps Learning of ML models. Review collected by and hosted on G2.com.

Two main disadvantages: access to foundation models (LLMs) is SLOW. Really slow. Direct access to other vendors LLMs, or even access to LLMs via Databricks in BDC is 3x faster (over API and SDK). AI Core does this way too slow for real production cases. Second thing: overall AI Core is complicated. UX is poor. AI Launchpad is getting better and better, but still it lags, there is no much access to logs, it is hard to learn navigation over AI Core (what object is doing what and when...) Review collected by and hosted on G2.com.