[BUG]

Environment

• Platform (select one):
• [ ] Anthropic API
• [X] AWS Bedrock
• [ ] Google Vertex AI
• [ ] Other: <!-- specify -->
• Claude CLI version: 0.2.69
• Operating System: macOS 14.4
• Terminal: iTerm2

Bug Description

I've been switching back and forth between using Anthropic API and AWS Bedrock. I've noticed that when using Anthropic there is 3.5 and 3.7 model use, but when using Bedrock there is only 3.7 model use. That results in a pretty large cost difference in terms of usage costs.

Steps to Reproduce

1. Use Ahtnropic API
2. Use claude code for tasks
3. Observe in Anthropic billing that there is usage of 3.5 and 3.7
4. Use AWS Bedrock
5. Use claude code for tasks
6. Observe in AWS that only 3.7 is being used.

Expected Behavior

It would be nice from a cost concern to have the tool use the cheaper model when possible.

Actual Behavior

Always used Claude 3.7

Additional Context

None