The API layer issues on Linux

Issues the maintainers labelled both platform:linux and area:api.

354 issues · 18 open · 336 resolved (95%) · first seen May 7, 2025

Is this getting better or worse?

This class of problem is converging. 26 new reports in the last 90 days vs 47 in the 90 before — -45%. The open backlog peaked at 85 in 2025-12 and sits at 18 today.

0 21 43 64 85 2025-052025-102026-032026-08
2025-05 — First reports2025-07 — Most new reports (43)2025-12 — Peak backlog (85 open)

Derived by counting each matching issue from its creation date until it closed. First-party computation over the anthropics/claude-code tracker — the trend is not published anywhere upstream.

Which builds report this

141 of these reports name the Claude Code build they were running, spanning 65 releases. Heaviest reporters:

Report volume, not a defect count: an issue can mention a build without that build having caused the failure. Versions come from the version string in the bug-report template, matched against real published tags.

Do these get fixed?

Median time to close is 30 days across 336 closures. Of the 336 closures with a recorded reason, 33% were closed as completed and 226 as not-planned or duplicate. The most recent completed fix landed in 2026-08.

Workarounds reported by the community

51 of these issues post a workaround someone says worked. The highest-engagement ones:

[BUG] Oversized image breaks conversation permanently - no way to recover without starting new chat
…retry without failed images, or client-side image resizing Workaround Currently, users must start a new conversation and manually reconstruct any important context.
Found in the description of #13480 · resolved
[BUG] No Response From API 2.1.181, 2.1.183 (constantly)
…-URL clients. Filed 69759 with the gateway-specific detail. Workaround that's holding for us: pin to 2.1.179 + so the background updater doesn't re-bump past it.
Found in the thread of #69358 · resolved
Claude Code: constantly seeing 'API Error (Request timed out...)' - Retrying in .. seconds.. (attempt x/10)'
…tep tasks, which can exceed the default API timeout limits. As a workaround, you can increase the API timeout by setting the API TIMEOUT MS environment variable before launching Claude Code. For example, you can use the…
Found in the thread of #2728 · resolved
[BUG] max_tokens must be greater than thinking.budget_tokens
workaround is possible by editing settings.json (in my case in ~/.claude/settings.json), by setting the "alwaysThinkingEnabled" to false.
Found in the thread of #8756 · still open
API Error 400: thinking blocks order error on first tool use after clean install
Another workaround is to use agents, it seems they are not affected by the bug.. It seems only the main claude get the error after using tools.
Found in the thread of #20692 · still open

Excerpts are quoted from the linked issue so you can judge relevance before opening it. Threads that only ask for a workaround, or report that there is none, are excluded. These are community suggestions, not official guidance from Anthropic — read the full thread before applying one.

Matching issues 1–100

Ranked by community engagement (reactions weighted over comments). Listing the top 300 of 354 matches; the statistics above cover all 354.

Related aggregations

  1. The API layer issues on macOS

    1,232 issues
  2. The API layer issues on Windows

    400 issues
  3. The API layer issues on VS Code

    64 issues
  4. Model behaviour issues on Linux

    1,695 issues
  5. Core CLI issues on Linux

    1,380 issues
  6. Terminal UI issues on Linux

    1,252 issues
  7. Security issues on Linux

    752 issues
  8. Tool calls issues on Linux

    676 issues