Normalize provider cache/usage token shapes
Cluster 70543 · agent. Data as of
Nothing open on GitHub. Fix merged in #85769 on 14 Aug 2026. Check it covers your case.
The short answer
- Known?
- Yes. 3 related items grouped here, 1 issue already closed.
- Fix event
- A fix was merged to main in PR #85769. Check it covers your case. No release data is tracked.
- Tracker
- The maintainers' tracker marks this
resolved, deciding PR #85769. That is separate from the fix event. - Still open
- Nothing open on GitHub.
These items address inconsistencies in how cache and usage token fields are normalized across multiple providers, specifically covering OpenAI cache write tokens and Qwen flat cached token formats. The work consolidates normalization logic to handle varying wire shapes from different provider APIs.
Open issues0
Open PRs0
Fixes on record
-
#85769 fix(usage): normalize every provider cache/usage wire shape in normalize_usage (5-PR consolidation)
files:
agent/usage_pricing.py,contributors/emails/[redacted:email],tests/agent/test_usage_pricing.py
1 PR closed without merging
From the close comments: merged via another change 1
-
#70522 fix(usage): normalize OpenAI cache_write_tokens and Qwen flat cached_tokens · merged via another change →
#85769,#66105,#70543
Closed issues · 1
- Closed#70543Usage accounting: OpenAI GPT-5.6+ cache writes and Qwen flat cache reads normalize to 0
How this cluster was grouped
-
Link evidence (
structural links throughout): every member is connected by structural evidence: shared changed files, or a closing/superseded reference. -
Pairwise link density
0.67: the share of item pairs that carry a link, counting a structural link in full and a relevance-only link at 0.36.1.00means every pair is linked.
Pick up a Lab task
Check whether a fix covers the open issues here, with your own agent. Bring back evidence for maintainers.
Open the LabAt a glance
- Subsystem
- agent
- Merged PRs
- #85769
- Closed issues
- 1
- Attempts
- 1 PR closed without merging
- Fix link checked
- 14 Aug 2026