Consistently error "No tool call found for function call output with call_id"
Open 💬 2 comments Opened Apr 13, 2026 by JounQin
💡 Likely answer: A maintainer (github-actions[bot], contributor)
responded on this thread — see the highlighted reply below.
What version of Codex CLI is running?
codex-cli 0.120.0
What subscription do you have?
Pro
Which model were you using?
gpt-5.3-codex
What platform is your computer?
Darwin 25.4.0 arm64 arm
What terminal emulator and version are you using (if applicable)?
VSCode
What issue are you seeing?
{"type":"error","status":400,"error":{
"message": "No tool call found for function call output with call_id call_Sp2iWQ8IWnaN7pmtKRxmn2ky.",
"type": "invalid_request_error",
"param": "input",
"code": null
}}What steps can reproduce the bug?
Uploaded thread: 019d8665-5784-7022-9945-c5da8210ab09
What is the expected behavior?
I just send a simple task request, so no error is expected.
Additional information
_No response_
2 Comments
Potential duplicates detected. Please review them and close your issue if it is a duplicate.
Powered by Codex Action
This tool call mismatch error is a serious reliability issue for AI coding tools. The system generates a function call with a call_id, but when it tries to provide the result, the backend can't find that call.
What's concerning is this happens consistently on simple tasks — it's not an edge case, it's a fundamental tool orchestration bug. When AI tools fail this way:
We've seen similar integration contract failures in AI-generated code: the AI calls an API with parameters that look correct but don't match the actual API contract, or it assumes a response format that the service doesn't provide.
This is why runtime validation of tool/API calls is essential. Static analysis can't catch contract mismatches that only appear at runtime — you need to validate that the call_id exists, that the parameters match the schema, and that the response can be parsed.
npx @opencodereview/cli scan . --sla L1