Comparison
Eldric vs AWS Bedrock.
Both give you AI capability. The difference is where it runs, who holds the data, and how you're billed. Here's the honest split.
| Eldric | AWS Bedrock | |
|---|---|---|
| Where it runs | Self-hosted, on your own hardware | AWS cloud, managed service |
| Data sovereignty | Complete — nothing leaves your network | Runs inside AWS |
| Offline / air-gap | Yes — no internet required | No |
| Cost model | By deployment size, flat | Per token / per API call |
| Backends | 30+ local & cloud model backends | AWS-hosted models |
| The stack | You own retrieval, agents, training, cluster | Managed API surface |
| Hardware | CPU or GPU, Pi to datacenter | AWS instances |
Honest guidance
When to use each.
Choose Bedrock
If you're all-in on AWS, don't have data-residency constraints, and prefer a fully managed service billed by usage.
Choose Eldric
If your data has to stay in-house, you want to run offline or air-gapped, or you'd rather own the whole stack and pay by deployment, not per token.