Direct provider calls
No proxy or added request latency.
Your OpenAI and Anthropic calls continue directly to the provider. SpendLens sends operational metadata after calls complete.
How it works
From AI spend to a model test your team can validate.
SpendLens connects provider spend to the workloads in your code, then identifies more cost-efficient models worth testing.
How SpendLens works
Three steps. No migration project.
Keep your provider and model calls. Add one Python decorator or wrap your Node.js client once, then see which workloads are driving spend.
1 Install2 Mark a workload3 See costs
Interactive · less than 60 seconds
No application rewrite
Add one small instrumentation step.
Add one Python decorator or wrap your Node.js client once, then mark the workloads that matter.