My FeedDiscussionsHashnode Enterprise
New
Sign in
Log inSign up
Learn more about Hashnode Headless CMSHashnode Headless CMS
Collaborate seamlessly with Hashnode Headless CMS for Enterprise.
Upgrade ✨Learn more
Greg Wilson

37 likes

·

1.5K reads

1 comment

Leo
Leo
Aug 19, 2024

Most of this isn't needed because you can use Anthropic on Bedrock through Anthropic's official Python package, avoiding all the annoying differences in the APIs, and avoiding having to manually handle all the boto3 stuff :)

The only thing you have to do is literally to replace "Anthropic" class client name with "AnthropicBedrock" and have the AWS ID/secret key/region in your env or provide them on initialization for that client. There are docs in GitHub README of the official Anthropic package.

·
·1 reply
Greg Wilson
Greg Wilson
Author
·Aug 20, 2024

I missed that! I just updated the blog post. For my use case, going direct to bedrock APIs still makes sense because I’m also interacting with S3, but for most use cases that don’t need other AWS services. This is a great option. Thank you!

·