InvokeAI

This release introduces significant new features including video generation, support for multiple new models (Krea, Ernie, Ideogram, Flux.2), multi-GPU and Intel XPU support, and FP8 precision for reduced VRAM usage. It also includes numerous performance improvements and enhancements to Canvas and workflow management.

This release introduces significant new features including video generation, support for multiple new AI models (Krea, Ernie, Ideogram, Flux.2), multi-GPU acceleration, and native Intel XPU support. It also brings performance improvements and enhancements to the Canvas and workflow management tools.

This is a security patch release addressing two critical vulnerabilities. It fixes a file overwrite issue allowing arbitrary code execution and a bug enabling connections to arbitrary networks.

This release introduces significant new features including video generation, support for several new models like Krea.2-Turbo and Ideogram 4, and multi-GPU capabilities. It also brings performance improvements and enhancements to the Canvas and workflow management tools.

This is a security patch release addressing critical vulnerabilities that could expose sensitive API keys and model information. It also includes fixes for less critical security issues and privilege escalation.

This patch release fixes a dictionary key error affecting Qwen Image and Qwen Image Edit models when utilizing the default Qwen 2.2 VL Encoder.

This maintenance release focuses on stability and bug fixes across various components, including UI responsiveness, model loading stability (especially with FP8 and Flux.2), and CI/CD pipeline improvements. Future major features like video generation are noted for version 6.14.0.

This maintenance release focuses on stability and bug fixes across the UI, model loading, and CI/CD pipelines. Key updates include support for PEFT named-adapter LoRAs and dependency upgrades for React, ROCm, and Transformers.

This release introduces extensive support for new locally-hosted models (Qwen, Anima) and externally-hosted models via API keys. Significant improvements were made across the Canvas, Workflow Editor, and Model Manager, including new LLM-powered prompt tools.

InvokeAI v6.13.0 RC3 introduces extensive support for new locally-hosted models (Qwen, Anima) and integration with external API providers like OpenAI and Google. This release also overhauls the workflow editor, adds LLM-powered prompt tools, and brings significant improvements to the Canvas and Model Manager.

InvokeAI v6.13.0 RC2 introduces extensive support for new locally-hosted models (Qwen, Anima) and externally-hosted providers (OpenAI, Gemini). This release also brings significant improvements to the Model Manager, Canvas, Workflow Editor, and adds LLM-powered prompt manipulation features.

InvokeAI v6.13.0 RC1 introduces extensive support for new locally-hosted models (Qwen, Anima) and externally-hosted providers (OpenAI, Gemini). This release also brings significant feature additions to the Canvas, Workflow Editor, and new LLM-powered prompt manipulation tools.

InvokeAI v6.12.0 introduces experimental multi-user support, enhanced model compatibility for Z-Image and FLUX.2, and significant new features for the Canvas, including text and gradient tools. This release also improves model and gallery management utilities and adds remote control capabilities via a new REST endpoint.

InvokeAI v6.12.0rc1 introduces experimental multi-user support, enhances compatibility with Z-Image and FLUX.2 models, and adds significant new features to the Canvas and Model Manager utilities. This release also provides new remote control capabilities via a REST endpoint.

InvokeAI 6.11.1 is a bugfix release addressing image generation quality issues, UI glitches, and model selection logic errors. It also includes new Russian translations.

This feature release introduces comprehensive support for the high-quality and fast FLUX.2 Klein (F2K) models, including new image editing capabilities. It also adds a Seed Variance Enhancer node for Z-Image Turbo to improve output diversity.

Resolved an issue with the ref_images metadata format for FLUX Kontext recall. The fix ensures correct parsing of reference image metadata by removing an unnecessary array wrapper.

InvokeAI v6.10.0 introduces major support for the Z-Image Turbo model family and several quality-of-life improvements, including workflow tagging, prompt weighting hotkeys, and a new hotkey editor. This release also adds VRAM management options and PBR map generation capabilities.

InvokeAI v6.10.0rc2 introduces significant backend support for the Z-Image-Turbo model family and adds several major quality-of-life features, including a Hotkey Editor, Workflow Tags, and bulk operations in the Model Manager.

InvokeAI v6.10.0rc1 introduces major backend support for the Z-Image-Turbo model family and significant UI enhancements, including a Hotkey Editor, Workflow Tagging, and bulk model operations. This release marks the first Community Edition update since the commercial venture closure.

This release significantly improves Invoke's Model Manager by refactoring model identification logic and normalizing the internal model storage directory structure. A key user-facing change is that previously unidentifiable models are now installed as 'Unknown' rather than being deleted.

This release significantly improves Invoke's Model Manager by refactoring model identification, introducing support for installing unknown models, and normalizing the internal model directory structure to a flat, UUID-based layout.

This release significantly improves Invoke's Model Manager by restructuring internal model storage to a flat, UUID-based directory and refactoring the model identification system. A key user-facing change is that previously unidentified models are now installed as 'Unknown' rather than being deleted.

This release significantly improves Invoke's Model Manager by refactoring model identification logic and changing the internal storage structure for Invoke-managed models to a flat, UUID-based directory system. A data migration is performed on first run to adapt existing models to the new structure.