From e29b70f4ede7988152bf27c2615e5bfa7e94c41d Mon Sep 17 00:00:00 2001 From: Workshop Participant Date: Wed, 9 Jul 2025 09:44:48 +0000 Subject: [PATCH] feat: updates documentation with cache token metric --- .../concepts/model-providers/custom_model_provider.md | 3 +++ docs/user-guide/observability-evaluation/metrics.md | 2 +- 2 files changed, 4 insertions(+), 1 deletion(-) diff --git a/docs/user-guide/concepts/model-providers/custom_model_provider.md b/docs/user-guide/concepts/model-providers/custom_model_provider.md index e7ff3aaf..26230257 100644 --- a/docs/user-guide/concepts/model-providers/custom_model_provider.md +++ b/docs/user-guide/concepts/model-providers/custom_model_provider.md @@ -222,6 +222,9 @@ Convert the event(s) returned by your model to the Strands Agents [StreamEvent]( "inputTokens": 234, # Number of tokens sent in the request to the model.. "outputTokens": 234, # Number of tokens that the model generated for the request. "totalTokens": 468 # Total number of tokens (input + output). + "cacheWriteInputTokens": 234 # Optional: Number of input tokens written to cache. + "cacheReadInputTokens": 0 # Optional: Number of input tokens read from cache. + } } ``` diff --git a/docs/user-guide/observability-evaluation/metrics.md b/docs/user-guide/observability-evaluation/metrics.md index 853f32e3..fcebd0a9 100644 --- a/docs/user-guide/observability-evaluation/metrics.md +++ b/docs/user-guide/observability-evaluation/metrics.md @@ -6,7 +6,7 @@ Metrics are essential for understanding agent performance, optimizing behavior, The Strands Agents SDK automatically tracks key metrics during agent execution: -- **Token usage**: Input tokens, output tokens, and total tokens consumed +- **Token usage**: Input tokens, output tokens, total tokens, and cache tokens consumed - **Performance metrics**: Latency and execution time measurements - **Tool usage**: Call counts, success rates, and execution times for each tool - **Event loop cycles**: Number of reasoning cycles and their durations