AI Playground
Everything in Cloud Playground plus Amazon Bedrock access — Claude, Titan, Llama, and Mistral models. Build RAG pipelines, AI agents, prompt flows, and computer vision apps with Rekognition. Perfect for AI/ML experimentation without long-term AWS commitments.
Amazon Bedrock Models Available
Claude (Anthropic)
Advanced reasoning, code generation, and analysis. Ideal for complex AI agents and multi-step workflows.
Titan (Amazon)
Text generation and embeddings. Great for search, summarization, and RAG pipelines.
Llama (Meta)
Open-weight large language model. Good for general-purpose text tasks and fine-tuning experiments.
Mistral
Efficient inference with strong coding and reasoning capabilities. Cost-effective for production workloads.
What You Can Build
RAG Pipelines
Build retrieval-augmented generation pipelines with Bedrock, S3, and DynamoDB. Index documents, embed with Titan, and query with Claude.
AI Agents
Create autonomous AI agents using Bedrock Agents or the Strands Agent SDK. Wire up tools, memory, and multi-step reasoning.
Prompt Engineering
Experiment with different models, temperatures, and system prompts. Compare outputs across Claude, Titan, Llama, and Mistral.
Computer Vision
Build image analysis apps with Amazon Rekognition. Detect objects, faces, text, and custom labels in images and video.
NLP & Text Analysis
Extract entities, sentiment, and key phrases with Amazon Comprehend. Build document processing pipelines.
Prompt Flows
Design multi-step prompt chains with Bedrock Prompt Flows. Connect models, tools, and conditional logic visually.
11 Services Included
Amazon Bedrock
Foundation model access — Claude, Titan, Llama, Mistral. Invoke models, create agents, and build prompt flows.
Claude
Anthropic's advanced reasoning model. Best for complex analysis, code generation, and multi-step agent workflows.
Titan
Amazon's native models for text generation and embeddings. Optimized for RAG and search use cases.
Rekognition
Image and video analysis — object detection, face recognition, text extraction, and custom labels.
Comprehend
NLP service for entity extraction, sentiment analysis, key phrases, and language detection.
SageMaker
Full ML platform — notebook instances, training jobs, model hosting, and MLOps pipelines.
Lambda
Serverless compute for AI inference endpoints, preprocessing, and event-driven ML pipelines.
S3
Object storage for training data, model artifacts, document corpora, and RAG knowledge bases.
DynamoDB
NoSQL database for conversation history, session state, agent memory, and vector metadata.
API Gateway
HTTP endpoints for AI-powered APIs. Wire up Lambda + Bedrock for production-style inference endpoints.
Step Functions
Orchestrate multi-step AI workflows — preprocessing, inference, post-processing, and human review.
Need More Time?
Extend your session without losing any work. Extensions apply instantly.
| Extension | Credits |
|---|---|
| +15 min | 4 credits |
| +30 min | 8 credits |
| +1 hour | 15 credits |
Build with AI on real AWS infrastructure
Launch a real AWS account with Bedrock, Claude, Titan, and 7 more services. Build RAG pipelines, AI agents, and ML experiments without the commitment.
Launch AI Playground →