AWS Bedrock Model Pricing
Multiple providers with different pricing. Smart routing across models delivers savings.
| Model | Input Cost | Output Cost | Provider | Best For |
|---|---|---|---|---|
| Claude 3.5 Sonnet | $3.00/1M | $15.00/1M | Anthropic | General purpose |
| Claude 3 Haiku | $0.25/1M | $1.25/1M | Anthropic | Simple tasks |
| Llama 3.1 70B | $0.99/1M | $0.99/1M | Meta | Open source |
| Llama 3.1 8B | $0.22/1M | $0.22/1M | Meta | Cost-effective |
| Mistral Large | $4.00/1M | $12.00/1M | Mistral | European AI |
| Amazon Titan Text | $0.50/1M | $0.65/1M | Amazon | AWS native |
Bedrock Optimization Features
Cross-Model Routing
Route between Claude, Llama, Titan, and Mistral based on task requirements and cost.
Provisioned Throughput
Optimize Provisioned Throughput allocation for predictable, high-volume workloads.
Response Caching
Cache similar queries across all Bedrock models for reduced API calls.
AWS Integration
Native integration with AWS Cost Explorer and CloudWatch for unified visibility.
Budget Controls
Set spending limits per model, application, or AWS account.
Cost Forecasting
ML-powered predictions for Bedrock spending based on usage trends.
Bedrock Use Cases
Multi-Model Strategy
Use Llama 3.1 8B for simple tasks, Claude Haiku for moderate, and Claude Sonnet for complex queries.
Knowledge Bases
Optimize Bedrock Knowledge Bases with smart embedding and retrieval strategies.
Agents for Bedrock
Optimize agent orchestration costs with intelligent model selection per step.