AWS Bedrock Pros and Cons

AWS Bedrock contains all LLM models which is helpful to choose the right model for the use cases. I built multiple Agents that helps under software development lifecycle and by using bedrock, I was able to achive the output in faster. Also the security features provided under Bedrock really helps to build chatbots and reduce error or hallucinations for text generation and virtual assistant use cases. Review collected by and hosted on G2.com.

Bedrock pricing model needs improvement. Few of the models are projected under AWS marketplace pricing. Bedrock not available in all regions and have to rely on US region for the same. Review collected by and hosted on G2.com.

AWS Bedrock makes it easy to bring generative AI into your business without getting bogged down in infrastructure details. It offers access to powerful AI models from top providers, like Anthropic and Amazon’s own Titan models, giving you flexibility to find the right fit. Bedrock is fully managed, so it scales as you grow and handles the heavy lifting on security and compliance. You can quickly build and deploy AI tools tailored to your needs, all within AWS, making it a great option for businesses ready to get real value from AI. Review collected by and hosted on G2.com.

AWS Bedrock can be costly, especially for small businesses, and it ties users tightly to the AWS ecosystem, limiting flexibility. Its complexity poses challenges for newcomers, and while it offers foundational models, it’s less adaptable than open-source options. Additionally, the documentation isn’t always user-friendly, making it harder to get up to speed quickly. Review collected by and hosted on G2.com.

I loved how well documented it is to use the APIs provided by AWS bedrock. They are providing models of their own as well as other companies like Meta as well. It is an API so you can use it very easily with the AWS platform as well. Review collected by and hosted on G2.com.

I feel like it is still a build in progress product with respect to availibilty in zones. One also might say it is costly but I feel like its a fair trade with the quality ans ease it offers. Review collected by and hosted on G2.com.

There are well trusted gen ai models provided by Amazon, meta and many more options to choose from. I appreciate their documentation as well. Easy API use from AWS CDK. Review collected by and hosted on G2.com.

In initial stages, it is a steep learning curve to go through, even though the documentation is good. The cost could be competitive with respect to other cloud service providers. Review collected by and hosted on G2.com.

it is easy to use powerful AI models without building everything from scratch. We can access different pre-trained models from top AI companies and quickly integrate them into our applications. This saves a lot of time and effort, especially for the businesses that want to use AI but don't have expertise in machine learning. it is also flexible, so we can choose models that fit our specific needs. Review collected by and hosted on G2.com.

For beginners, AWS Bedrock can be little bit hard to set up. And always AWS services are little bit coslty. Review collected by and hosted on G2.com.

while working as a genAI engineer, I just have to use the LLM model from bedrock, the availibity of llm mode is high and user doesn't need to worry about configuration too much, architecture and configuration alreay taken care by aws bedrock. Review collected by and hosted on G2.com.

If I want to use customised model in aws , there is limitation and also aws bedrock is not available in London region. It's availibity need to increased. Review collected by and hosted on G2.com.

I like the fact that a lot of pre-trained models are available which can be quickly integrated. Also, if a company is already in the AWS ecosystem, it is easy to integrate with Bedrock and continue the processing pipeline by integrating it with other managed AWS services.

Budget alarms allow to scale it to what is needed without frequently bottlenecking the application.

AWS is a trusted and reliable cloud provider hence furthermore solidifying the use case of using Bedrock for deploying GenAI. Review collected by and hosted on G2.com.

If a product is not already in the AWS ecosystem, or if a company prefers to deploy on bare metal, then using Bedrock can cause vendor lockin.

For very niche scenarios, a lot of tweaking is required. Fully training model from scratch can be a costly process. Review collected by and hosted on G2.com.

Bedrock is a great way to integrate AI services into your existing amazon infrastructure Review collected by and hosted on G2.com.

As new models and features become available in the third party LLM, they are often not available on Bedrock right away Review collected by and hosted on G2.com.

Amazon Bedrock offers single API with access to high perorming foundation models like Stability AI, Mistral AI and many other providers. Instead of playing around each API seperately, you can integrate and fine tune your models easily with AWS Bedrock Review collected by and hosted on G2.com.

AWS Bedrock can be little expensive, compared to using the foundation models directly, but saves a lot of effort on managing models indivdually. Review collected by and hosted on G2.com.

The best part of using services of AWS Bedrock is availablity of different models. Since its fully managed, and we using Llama, its a total saviour to use a service which manages everything. Review collected by and hosted on G2.com.

The finetune is not aavaialable. For that againt we have to use sagemaker. The back and forth of using different service gives a little friction and everything cant be done in one place. Review collected by and hosted on G2.com.