[Bug] Model refusing requests for legitimate security application

Open 💬 0 comments Opened Jul 3, 2026 by ghost

Bug Description
What in the WORLD is up with this model blocker.. I am building an absolutely BASIC app with security features for the user to use to protect themselves at the most basic level and all of a sudden EVERY message is getting flagged? it's literally a digital pen handwriting capture app. Fable5 already burned 4.5m+ tokens on a deep research I didn't ask for that it did immediately after on a targeted search for roughly 6k tokens... is this what my $250/mo subscription is buying me? blocked progress and intentionally wasting my tokens?

Environment Info

  • Platform: linux
  • Terminal: xterm-256color
  • Version: 2.1.197
  • Feedback ID: ebb7314c-1504-4869-8273-9655ab653c17

Errors

[{"error":"Error: 413 {\"error\":{\"type\":\"request_too_large\",\"message\":\"Request exceeds the maximum size\"}}\n    at generate (/$bunfs/root/src/entrypoints/cli.js:12:71094)\n    at makeRequest (/$bunfs/root/src/entrypoints/cli.js:52:7688)\n    at processTicksAndRejections (native:7:39)","timestamp":"2026-07-02T00:30:07.925Z"}]

View original on GitHub ↗