ton/pi_harness
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat(ai): add reasoning token counts to Usage

Browse Source 
Add optional reasoning?: number to Usage as a subset of output. Populate
for Anthropic (output_tokens_details.thinking_tokens), OpenAI
Responses/Codex/Azure (output_tokens_details.reasoning_tokens), OpenAI
Completions (completion_tokens_details.reasoning_tokens), and Google
Generative AI / Vertex (thoughtsTokenCount). Bedrock Converse and Mistral
do not return a reasoning breakdown, so they stay unset.

closes #6057
This commit is contained in:
Mario Zechner
2026-06-25 10:34:47 +02:00
parent 5c76ae407d
commit d7868b0998
7 changed files with 23 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
packages/ai/CHANGELOG.md
+4
View File
@@ -2,6 +2,10 @@
## [Unreleased]
### Added
- Added an optional `reasoning` field to `Usage` reporting reasoning/thinking token counts as a subset of `output`. Populated for Anthropic (`output_tokens_details.thinking_tokens`), OpenAI Responses/Codex/Azure (`output_tokens_details.reasoning_tokens`), OpenAI Completions (`completion_tokens_details.reasoning_tokens`), and Google Generative AI / Vertex (`thoughtsTokenCount`). Bedrock Converse and Mistral are not populated because those APIs do not return a reasoning token breakdown ([#6057](https://github.com/earendil-works/pi/issues/6057)).
### Fixed
- Fixed retry classification for provider errors that explicitly tell callers to retry the request ([#6019](https://github.com/earendil-works/pi/issues/6019)).