Skip to content

/goal completion should show token usage breakdown, not only combined usage #28083

Open

Description

@nelsonjchen

What variant of Codex are you using?

Codex CLI and Codex Desktop/App, using Goal mode (/goal)

What feature would you like to see?

Summary

/goal currently reports combined usage when a goal completes. I would like the goal completion message to also show a token usage breakdown.

For example, instead of only showing an effective/combined total, the completion receipt could include:

  • input tokens
  • cached input tokens
  • output tokens
  • reasoning output tokens
  • effective/combined usage

Why this matters

Goal mode is often used for long-running autonomous work. When a goal finishes, a single combined usage number is not enough to understand where the usage went.

A breakdown would help users tell whether a goal was expensive because of:

  • large input/context
  • poor prompt-cache reuse
  • high reasoning output
  • large final/output responses
  • repeated continuation turns
  • compaction or retry overhead

This would make /goal completion receipts much more useful for debugging unexpected usage and comparing long-running goal sessions.

Expected behavior

When a goal completes, Codex should show something like:

Goal complete

Usage:
- Effective usage: 1,234,567
- Input tokens: 900,000
- Cached input tokens: 7,500,000
- Output tokens: 120,000
- Reasoning output tokens: 45,000

Additional information

I searched for existing issues and found adjacent reports about usage visibility and goal token behavior, but not this exact request.

Related but not duplicate:

This request is specifically about the /goal completion receipt showing a usage breakdown rather than only combined usage.

Activity

  1. added
    CLIIssues related to the Codex CLI
    appIssues related to the Codex desktop app
    rate-limitsIssues related to rate limits, quotas, and token usage reporting
    on Jun 13, 2026
  2. zinman commented on Jul 22, 2026

    @zinman

    Adding another user-impact data point in support of this request.

    In Codex Desktop/App chat, a user asked for the equivalent of Claude Code's /usage report: per-session model usage, input/output/cached/reasoning tokens, wall/API time, limit reset information, and USD cost or a clearly labeled estimate. From inside the app session, the assistant could only report that no active goal usage report was available and that billing/cost data was not exposed.

    It would be very useful if Desktop/App tasks exposed a session/task receipt similar to:

    • model(s) used
    • input tokens
    • cached input tokens
    • output tokens
    • reasoning output tokens
    • effective/combined usage
    • wall time and API time where available
    • estimated cost or explicit "cost unavailable" reason
    • relevant plan/limit reset status

    This should ideally be available for normal tasks, goal completion receipts, and historical task inspection. The main value is not just debugging token burn; it lets users understand whether a task was expensive because of context size, cache behavior, reasoning effort, tool loops, or output volume.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    CLIIssues related to the Codex CLIappIssues related to the Codex desktop appenhancementNew feature or requestrate-limitsIssues related to rate limits, quotas, and token usage reporting

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions