This is a Release Candidate. We strongly suggest backing up your database before testing to prevent data loss in case of any issues.

Please let us know if you run into anything unexpected.

What's Changed in RC6 from RC5

🎉 Barring any major issues, this will be the last RC before v4.0.0! 🎉

• Updater removed - use the installer to update
• Installation docs rewritten
• FAQ updated
• Iterate nodes now iterate in order - thanks @cgi-joe!
• Fix soft lock when model install fails
• Autoimport removed - use Scan Folder instead
• Sane workflow sorting - thanks @clsn!
• Numerous minor fixes and enhancements

What's New in 4.0

Simplified Installation, Updating and Configuration

We've simplified and streamlined installation, making it much faster and reliable:

• No configuration script - all config is in invokeai.yaml
• No model installs during installation
  • Internal "core" models are downloaded only when they are first requested
  • Generation models are installed with the Model Manager UI
• No updater - use the installer instead

New Model Manager

The model manager is rewritten in v4.0.0, both frontend and backend. This builds a foundation for future model architectures and brings some exciting new user-facing features:

• All model installation happens via the UI (no configure script)
• Queued model downloads
• Per-model preview images
• Per-model default settings - choose a model’s default VAE, Scheduler, CFG Scale, etc
• User-defined trigger phrases for concepts/LoRAs and models - access by typing the < key in any prompt box
• API key support for model marketplaces
• Autoimport removed - use Scan Folder instead

Model Hashing

When you first run v4.0.0, it may take a few minutes to start up as it does a one-time hash of all of your model files.

Do not panic.

Hashes provide a stable identifier for a model that is the same across every platform.

If you don’t care about this, you can press Ctrl+C to interrupt the process and disable hashing by setting hashing_algorithm: random setting in invokeai.yaml.

Canvas Improvements

The canvas uses a new method for compositing called gradient denoising. This eliminates the need for multiple “passes”, greatly reducing generation time on the canvas. This method also provides substantially improved visual coherence between the masked regions and the rest of the image.

The compositing settings on canvas allow for control over the gradient denoising process.

Major research & experimentation for this novel denoising implementation was led by @dunkeroni, and @blessedcoolant was responsible for managing integration into the canvas UI.

Fixes and Enhancements

Many small bug fixes, resolved papercuts, and warm fuzzies. Shouting out some notable goodies from the community:

• Bulk downloads (download a selection of images or a whole board) @StefanTobler
• Canvas Brush Size Scroll can now be inverted @joshistoast
• Images in the Canvas Staging Area can now be discarded individually @joshistoast
• Numerous fixes and UI enhancements @joshistoast
• Numerous greybeard node things @dunkeroni
• Iterate nodes now iterate in order @cgi-joe
• Sane workflow sorting @clsn
• Image dimensions overlay in the gallery @rohinish404
• Localization fixes @rohinish404
• New translations B N, @Harvester62, @Pfannkuchensack, @Bethanielle, @Vasyanator, @GGSSKK, & @Sufi2425
• Updated torch and diffusers deps @Malrama
• Docs updates @skunkworxdark, @gogurtenjoyer
• LoRA probe fix @skunkworxdark

Bonus: Invoke Training (Beta)

As of v4.0.0, all references to training in the core invoke script now point to the Invoke Training Repo. Invoke Training offers a simple user interface for:

• Textual Inversion Training
• LoRA Training
• Dreambooth Training
• Pivotal Tuning Training

Learn more on the Invoke Training repo.

Installation and Upgrading

To install or upgrade to version 4.0, download the zip file from the release notes ("Assets" section), unzip it, and follow the installation instructions. For upgrades, select the same installation location.

💾 Download Installer

💻 Developer Changes

v4.0.0 is versioned as a major release due to breaking changes:

• The internal nodes API has been refactored to provide a stable public API. Node authors should review the migration guide.
• The internal graph execution engine is drastically simplified, resulting in more efficient and performant processing. This carries on from the changes in v3.6.0 in which graphs are no longer stored in the database.

Contributing

As a community-supported project, we rely on volunteers and enthusiasts for continued innovation and polish. Everything from minor documentation fixes to major feature additions are welcome. To get started as a contributor, please refer to How to Contribute or reach out in #dev-chat on Discord!

What's Changed

• fix:Update probe.py - Case-sensitive typo on .LoRA by @skunkworxdark in https://github.com/invoke-ai/InvokeAI/pull/6031
• allow deletion of symlinked models in models dir by @lstein in https://github.com/invoke-ai/InvokeAI/pull/6036
• fix(mm): default settings pydantic error by @psychedelicious in https://github.com/invoke-ai/InvokeAI/pull/6045
• fix(ui): load default workflows mutation error by @psychedelicious in https://github.com/invoke-ai/InvokeAI/pull/6050
• tidy: launcher by @psychedelicious in https://github.com/invoke-ai/InvokeAI/pull/6046
• fixes ui localization and settings typo by @rohinish404 in https://github.com/invoke-ai/InvokeAI/pull/6033
• Change GraphExecutionState's _get_next_node method to prioritize Iterate Nodes by @cgi-joe in https://github.com/invoke-ai/InvokeAI/pull/6032
• Update transformers and diffusers by @Malrama in https://github.com/invoke-ai/InvokeAI/pull/6037
• feat(installer): remove updater & support installing from wheel by @psychedelicious in https://github.com/invoke-ai/InvokeAI/pull/6054
• docs: new installation docs by @psychedelicious in https://github.com/invoke-ai/InvokeAI/pull/6056
• fix(mm): prevent installer softlock when an unhandled error occurs by @psychedelicious in https://github.com/invoke-ai/InvokeAI/pull/6048
• feat(mm): remove MALLOC_MMAP_THRESHOLD_ env var by @psychedelicious in https://github.com/invoke-ai/InvokeAI/pull/6059
• fix(mm): handle integer state dict keys in probe by @psychedelicious in https://github.com/invoke-ai/InvokeAI/pull/6051
• fix(ui): image not getting selected when clicked on image-badge by @rohinish404 in https://github.com/invoke-ai/InvokeAI/pull/6057
• fix(mm): handle depth and inpaint ckpt conversion by @psychedelicious in https://github.com/invoke-ai/InvokeAI/pull/6064
• docs: remove handful of deprecated docs by @psychedelicious in https://github.com/invoke-ai/InvokeAI/pull/6063
• feat: display torch device on startup by @psychedelicious in https://github.com/invoke-ai/InvokeAI/pull/6062
• feat(mm): remove hf token handling by @psychedelicious in https://github.com/invoke-ai/InvokeAI/pull/6061
• Check cuDNN version compatibility on startup by @RyanJDick in https://github.com/invoke-ai/InvokeAI/pull/6052
• Flip default ordering of workflow library; #5477 by @clsn in https://github.com/invoke-ai/InvokeAI/pull/6066
• feat(ui): when using a control image's dimensions, fit to the current model's optimal dimensions by @psychedelicious in https://github.com/invoke-ai/InvokeAI/pull/6068
• chore(ui): bump deps, cleanup by @psychedelicious in https://github.com/invoke-ai/InvokeAI/pull/6070
• feat(mm): remove autoimport, revised startup model scanning by @psychedelicious in https://github.com/invoke-ai/InvokeAI/pull/6060
• chore: v4.0.0rc6 by @psychedelicious in https://github.com/invoke-ai/InvokeAI/pull/6071
• fix(installer): restore install successful message by @psychedelicious in https://github.com/invoke-ai/InvokeAI/pull/6073
• chore: ruff by @psychedelicious in https://github.com/invoke-ai/InvokeAI/pull/6074

New Contributors

• @cgi-joe made their first contribution in https://github.com/invoke-ai/InvokeAI/pull/6032
• @clsn made their first contribution in https://github.com/invoke-ai/InvokeAI/pull/6066

Full Changelog: https://github.com/invoke-ai/InvokeAI/compare/v4.0.0rc5...v4.0.0rc6