Feature: Show per-request token counts in CLI
Resolved 💬 3 comments Opened Feb 28, 2026 by sareenb Closed Mar 28, 2026
Feature Request
Display input and output token counts for each request/response cycle in the Claude Code CLI.
Motivation
Currently, there is no way to see how many tokens are being consumed per request in the CLI. The only usage visibility is the aggregate percentage on claude.ai, which doesn't help users understand:
- Which prompts are token-heavy
- How context window usage grows over a session
- Whether they are approaching session limits
Proposed Behavior
After each response, optionally show a token usage summary, for example:
[Tokens: 1,234 in / 567 out | Session: 12,450 total]
Or via a flag like --show-tokens / a setting in config.
Alternatives Considered
- Checking claude.ai usage dashboard — only shows aggregate %, not per-request breakdown
- Using the raw API directly — not practical for CLI users
Environment
- Tool: Claude Code CLI
This issue has 3 comments on GitHub. Read the full discussion on GitHub ↗