Interesting setup
do you have any benchmarks on: - token usage over time - failures/retry rates
would be great to see how it behaves in production
Interesting setup
do you have any benchmarks on: - token usage over time - failures/retry rates
would be great to see how it behaves in production