Issues

• [bug]: webv2 resubmits a run indefinitely after a 500 on enqueue; the run is hidden from the Queue panel and survives reloads

  invoke-ai/InvokeAI#9854 · Filter by author unmarco opened on Oct 14, 2026

• [bug]: Enqueue returns 500 when a Collect node feeds a Union[list[X], X] / Optional[list[X]] input with the wrong item type

  invoke-ai/InvokeAI#9852 · Filter by author unmarco opened on Oct 14, 2026

• [Enhancement]: Different image formats and better compression

  enhancement
  invoke-ai/InvokeAI#9850 · Filter by author Deniryer opened on Oct 14, 2026

• [bug]: sibling worker can consume a GPU borrow-release wake-up

  invoke-ai/InvokeAI#9849 · Filter by author wunianze666-netizen opened on Oct 14, 2026

• [bug]: Qwen3-VL-32B-Int8 Text Encoder isn't recognized as MiniMax-H3 TE component

  bug
  invoke-ai/InvokeAI#9846 · Filter by author NiceDragon8066 opened on Oct 14, 2026
  1

• [bug]: Interrupted model download restarts from the beginning for download retry option

  bug
  invoke-ai/InvokeAI#9841 · Filter by author NiceDragon8066 opened on Oct 13, 2026
  1

• [bug]: Wan-2.2 A14B High and Low Noise models cannot be loaded simultaneously. Standard workflows are missing.

  bug
  invoke-ai/InvokeAI#9840 · Filter by author NiceDragon8066 opened on Oct 13, 2026

• [enhancement]: Need support for NVFP4 text encoder Qwen3-VL-32B for MiniMax-H3

  enhancement
  invoke-ai/InvokeAI#9839 · Filter by author NiceDragon8066 opened on Oct 13, 2026

• [bug]: XLabs FLUX ControlNet v3 checkpoints are installed as FLUX main models

  bugmodel manager
  invoke-ai/InvokeAI#9832 · Filter by author Pfannkuchensack opened on Oct 12, 2026

• [bug]: Wan 2.2 A14B at 720p does not fit 24 GB (no working-memory estimate); SageAttention makes the overflow slower

  bug
  invoke-ai/InvokeAI#9831 · Filter by author Pfannkuchensack opened on Oct 12, 2026

• [bug]: "Retry install" in model manager fails for gated models requiring a Hugging Face token after adding it

  bug
  invoke-ai/InvokeAI#9829 · Filter by author AuralVertigo opened on Oct 12, 2026

• [enhancement]: Bump PyPatchMatch to latest

  enhancement
  invoke-ai/InvokeAI#9826 · Filter by author iwr-redmond opened on Oct 12, 2026

• [bug]: Long node chains into an If branch raise RecursionError

  bug
  invoke-ai/InvokeAI#9821 · Filter by author lstein opened on Oct 11, 2026
  

• [bug]: Per-item If over a collected sibling-iterator zip is rejected as UnsupportedIfFramesError

  bug
  invoke-ai/InvokeAI#9820 · Filter by author lstein opened on Oct 11, 2026
  

• [bug]: For loop over an empty Iterate item fails with NodeInputError

  bug
  invoke-ai/InvokeAI#9819 · Filter by author lstein opened on Oct 11, 2026
  

• [bug]: If inside a For body cannot select root-frame branch values (hang on main, KeyError with #9789)

  bug
  invoke-ai/InvokeAI#9818 · Filter by author lstein opened on Oct 11, 2026
  

• [enhancement]: Adjustment layers: Vibrance/Saturation and Clarity/Dehaze

  enhancement
  invoke-ai/InvokeAI#9813 · Filter by author shanedk opened on Oct 11, 2026

• [enhancement]: Add "Keep Input Language" Option to Prompt Expander (with Auto-Translation on Generation)

  enhancement
  invoke-ai/InvokeAI#9812 · Filter by author baramofme opened on Oct 11, 2026
  1

• [enhancement]: Load torchao-quantized single-file transformers (unsloth FP8)

  enhancement
  invoke-ai/InvokeAI#9798 · Filter by author Pfannkuchensack opened on Oct 10, 2026

• In-place install always on when installing from configured model folder

  7.0.0bug
  Bug

  invoke-ai/InvokeAI#9796 · Filter by author joshistoast opened on Oct 10, 2026

• In-progress preview reactivates on invocation after disabling

  7.0.0
  Bug

  invoke-ai/InvokeAI#9795 · Filter by author joshistoast opened on Oct 10, 2026

• [bug]: FLUX.2 and Z-Image scheduler paths ignore the model sigma schedule and denoising_end; Heun progress stops short

  bug
  invoke-ai/InvokeAI#9792 · Filter by author Pfannkuchensack opened on Oct 10, 2026

• [bug]: LCM scheduler is not reproducible with a fixed seed (FLUX.1, FLUX.2, Z-Image)

  bug
  invoke-ai/InvokeAI#9791 · Filter by author Pfannkuchensack opened on Oct 10, 2026

• [bug]: FLUX Denoise crashes when continuing from latents with add_noise=false (two-stage chain)

  bug
  invoke-ai/InvokeAI#9790 · Filter by author Pfannkuchensack opened on Oct 10, 2026

• [enhancement]: 7.0.0-alpha.1-post - UX/UI make (list) element checkboxes toggle with element body when there is enough space or no other event attached

  enhancement
  invoke-ai/InvokeAI#9785 · Filter by author Juarrow opened on Oct 10, 2026