Claude Agents SDK and the 5-hour limit: what is on the record
Users report that traffic labeled as the Claude Agent SDK / headless claude -p uses more of the Claude Pro or Max five-hour usage window per token than the interactive Claude Code app. This page lists only what appears in Anthropic's own GitHub repository issue tracker and Anthropic help pages. We do not treat third-party posts on X as sources. Where Anthropic has not published a statement, we write not stated.
Short answer
- An open bug report on the official
anthropics/claude-coderepository (issue #97074, opened 2026-09-25) describes headlessclaude -p/sdk-cliusing about 1.8× more of the five-hour window per token than interactiveclifor the same workload. - Anthropic help says Agent SDK usage still draws from the subscription's usage limits. It does not say the five-hour window is weighted differently by entrypoint.
- As of this page's check date, we found no Anthropic docs changelog entry, status-page incident, or named Anthropic staff post that confirms the gap as a bug or as intended policy.
- Whether the difference is a bug or intentional metering: not stated by Anthropic on the pages checked here.
What the anthropics GitHub issue says
On a Max subscription, the same workload uses about 1.8× more of the 5-hour window when it runs headless (claude -p --input-format stream-json, recorded as entrypoint: "sdk-cli" in the transcripts) than when it runs in the interactive TUI (entrypoint: "cli"). Token usage is the same in both modes.
The same tokens on the same model should use the same share of the 5-hour window, whether they come from claude -p or from the interactive TUI. If headless usage is intentionally weighted differently, it should be documented.
The issue remains open as of the snapshot date above. Comments on that thread include additional reproductions by other accounts; those are user reports on Anthropic's tracker, not Anthropic staff confirmations.
What Anthropic's help center says
still draw from your subscription limits
That article (updates dated 2026-06-15 and 2026-10-07 in the snapshot) states that Agent SDK / claude -p usage counts against the same subscription limits when you are signed in with your Claude plan. A different weight for sdk-cli vs cli: not stated on that page.
Status page
The Claude status page (https://status.claude.com/, checked 2026-10-25) is for service availability. A status incident that names Agent SDK five-hour metering: not stated in the material we quote for limits guides. Live feed mirror: Is Claude down?.
What the sources don't say
- An Anthropic blog post, changelog, or named staff account confirming the gap: not stated in the sources above.
- Whether weekly limits are affected: the GitHub issue discusses the five-hour window; Anthropic's position: not stated.
- Any official workaround from Anthropic: not stated. Community workarounds on the GitHub thread are not Anthropic documentation.
Sources
- https://github.com/anthropics/claude-code/issues/97074 (checked 2026-10-27)
- https://support.claude.com/en/articles/15036540-use-the-claude-agent-sdk-with-your-claude-plan (checked 2026-10-27)
See also: Claude vs Codex limits, side by side · Claude limits draining faster? What's confirmed · How Claude's weekly usage limits work · Claude status tracker · AI coding plans. Something wrong? Dispute it.