InvokeAI · Issues· 371 open

[enhancement]: Image graph does not contain random seeds used to generate an image (Could be a bug)

[bug]: Crashing Krea2 non-standard checkpoints after recent update.

[bug]: repo_name error when using starter SDQN model

[bug]: Official FLUX.2 Klein Qwen3 8B Encoder installs as Unknown; metadata update fails on macOS

[enhancement]: Default to Flashpack for v7

[bug]: Every drag makes the window freeze until refresh since the last update

[bug]: Segment Anything V2 fails every time

[bug]: InvokeAI AMD 96 GB Unified Memory / Windows ROCm Issue

[enhancement]: Optional acceleration for non-distilled models

[bug]: Flux (GGUF) broken on MPS, updating PyTorch fixes

[bug]: "i" hotkey on canvas no longer activates the eyedropper tool

[bug]: Black images don't work for Inpaint Mask, Images less than 600px can't be used for Flux.2 Ref image

[bug]: Potential Same CUDA Graphs Kernel Path Failure Problem as Llama CPP? (Proposed workaround included)

Documentation: amdgpu lowered lockup_timeout in kernel 7.x affects InvokeAI

[enhancement]: Add video duration in seconds with automatic frame count calculation