AI API receipt · BerriAI/litellm
BerriAI/litellm#45702 · upstream bug reproduced by running the issue's own reproduction
Badgr cannot tell this response is wrong
With one deployment limited to tpm=1, six /v1/messages calls all answered HTTP 200 and three were served by the TPM=1 deployment; the same limit returned 429 on /v1/chat/completions (reproduced on a live LiteLLM 1.104.2 proxy).
a 200 with a normal answer is a valid response; Badgr cannot know the deployment's TPM limit was already exceeded. The receipt therefore records success. This is a limit of what the contract can prove, shown here so it is not mistaken for protection.
The output the bug produces, taken from the issue’s reproduction and fed to Badgr’s validators. Not a live provider call.
text: "anthropic-leg" tool_calls: none
Written by Badgr’s receipt ledger and read back from GET /v1/receipts/replay-litellm-45702-messages-tpm-not-enforced.
{
"request_id": "replay-litellm-45702-messages-tpm-not-enforced",
"status": "success",
"endpoint": "/v1/messages",
"stream": false,
"failure_stage": null,
"error_code": null,
"error_message": null,
"retryable": null,
"provider_mode": "replay",
"routing_decision": "replayed_upstream_output",
"created_at": "1791679612.9811862"
}Recorded 2026-10-11 in Badgr’s local development environment. Replay cases and verdicts are pinned by backend/test_ai_api_replay_cases.py.
← All AI API receipts