|
At the moment, we’re asserting a brand new console expertise in Amazon Bedrock so that you can experiment, iterate, and scale with the most recent AI fashions on Amazon Bedrock’s next-generation inference engine constructed for prime efficiency, reliability, and safety. This console has a refreshed workflow optimized for bedrock-mantle endpoint, which helps the most recent GPT, Claude, and open-weight fashions with the OpenAI Responses API, OpenAI Chat Completions API, and the Anthropic Messages API.
The brand new console expertise makes it easy to search out the appropriate mannequin and transfer rapidly from analysis to manufacturing.
- New mannequin card – You may browse the total mannequin catalog, evaluate them facet by facet on capabilities, modality assist, context window, and relevant service quotas in a single view, eradicating the necessity to sew collectively documentation, and restrict calculators.
- Venture-based work – You can also make a mission to run evaluations and evaluate utilization insights in a single streamlined workflow that mirrors the lifecycle of constructing a generative AI utility.
- Dwell documentation – You should utilize project-aware dwell documentation: code samples, SDK snippets, and API references are mechanically prefilled along with your mission variables. You may copy a snippet straight from the console into your utility and run it with out modification.
get began
You may attempt a brand new expertise by selecting Strive the Bedrock Mantle Console from throughout the Amazon Bedrock console, or through the use of the new console hyperlink straight.
You could find a project-based dashboard to indicate inference requests and error by vary of latest dates, just lately used fashions, and the mission record. You may create a mission, assign fashions, configure API keys, and begin making inference requests in minutes.
A brand new mannequin catalog reveals the most recent GPT, Claude, and open-weight fashions which might be supported on the bedrock-mantle engine. You may see the main points of options, tokens, pricing, enter/output, pricing data, and Regional availability. You may as well evaluate as much as 3 fashions in a single view.
Whenever you select the mission dashboard, you possibly can see the fashions used within the mission, the distribution of your token utilization comparable to whole token utilization, token utilization per minute, inference requests per minute, and tokens per inference request. This may inform your mannequin choice, immediate optimization, and workload consistency selections.
You may choose as much as 3 fashions to begin evaluating to match responses facet by facet with the identical immediate.
To construct your utility within the mission, select Getting began. You may migrate present code, construct a brand new app with the Anthropic or OpenAI SDK, or join an AI coding assistant to Bedrock.
Select the API & SDK, your SDK (both Anthropic or OpenAI), your most popular programming language, and your authentication technique. It reveals your atmosphere code to run these in your terminal for a fast check, or save to a .env file to your utility. You may as well ship your first request with pattern code snippets to confirm your setup.
Whenever you select Shoppers, you possibly can choose the AI coding agent supply comparable to Claude Code, Cline, Codex, Cursor, or OpenCode that you just wish to hook up with the bedrock-mantle engine. It offers directions on how you can set up the AI agent, use your AWS IAM credentials or use a Bedrock API key, set atmosphere variables, and route requests from every AI agent by way of Bedrock.
To find out about Anthropic- and OpenAI-compatible APIs, select Dwell API docs. You may select Anthropic API Protocol for entry to Claude mannequin options just like the Messages API or OpenAI API Protocol for entry to options like Responses API.
For instance, whenever you select OpenAI Response API, it retrieves a mannequin response with the given mannequin ID. These API references are mechanically prefilled with the mission’s chosen mannequin ID, Area, bedrock-mantle endpoint URL, and API key reference, and so they replace in place as you alter fashions or settings.
You may as well select the prevailing Bedrock console to handle fully-managed options comparable to Brokers, Information Bases, Guardrails, fine-tuning, or the InvokeModel and Converse APIs to run on the bedrock-runtime endpoint.
Now out there
The brand new console expertise is on the market in all AWS Areas the place the bedrock-mantle endpoint is obtainable: US East (N. Virginia, Ohio), US West (Oregon), Asia Pacific (Jakarta, Mumbai, Sydney, Tokyo), Europe (Frankfurt, Eire, London, Milan, Stockholm), and South America (São Paulo). Test the full record of Areas for future updates.
Give the brand new console expertise a attempt within the new Amazon Bedrock console and ship suggestions to AWS re:Put up for Amazon Bedrock or by way of your ordinary AWS Help contacts.
— Channy