mflux releases

🎨 New Model Support

• FIBO Edit: Add image-editing support for the FIBO model family.
• FIBO Edit remove-background workflow: Support the dedicated remove-background edit path for FIBO.

✨ Improvements

• Training image scaling: Scale training images by area rather than longest side for more consistent preprocessing.
• MLX 0.31.x: Allow MLX 0.31.x in dependency ranges.
• FLUX.2 LoRA mapping: Expand LoRA key mapping coverage for FLUX.2.

🐛 Bug Fixes

• Training optimizer state: Evaluate optimizer state after each training step as intended.
• Local tokenizer loading: Fix loading tokenizers from local paths.
• Dynamic-resolution image edit: Restore correct behavior for image edit when using dynamic resolution.

👩‍💻 Contributors

• @filipstrand
• @icelaglace
• @TheOrsa
• @waldheinz

✨ Improvements

• Broader LoRA compatibility for FLUX.2 and Z-Image: Expand LoRA mapping coverage so more adapter key layouts resolve cleanly for FLUX.2 and Z-Image models.

👩‍💻 Contributors

• @filipstrand

✨ Improvements

• Local-model LoRA training: Allow LoRA training to work when the base model is supplied from a local path, including the FLUX.2 and Z-Image training adapters.

📝 Documentation

• Distilled-model step defaults: Clarify CLI guidance so examples prefer model default inference steps unless the user intentionally overrides them.

👩‍💻 Contributors

• @waldheinz

🎨 New Model Support

• FIBO-Lite support: Add support for the FIBO-Lite model variant.

🐛 Bug Fixes

• FLUX.2 edit downsampling extents: Fix downsampling in FLUX.2 edit paths so image extents are preserved.

👩‍💻 Contributors

• @filipstrand

✨ Improvements

• SeedVR2 7B support: Add support for the SeedVR2 7B upscaler variant.
• Qwen-Image parity with diffusers: Align Qwen-Image behavior more closely with the diffusers reference implementation.
• FIBO scheduler default: Default FIBO generate_image to flow_match_euler_discrete.

🧰 DX & Maintenance

• Repo tooling cleanup: Remove unused Cursor command wrappers from the repository.
• SeedVR2 7B test coverage: Add image test support for the new SeedVR2 7B path.

👩‍💻 Contributors

• @ciaranbor
• @icelaglace
• @filipstrand

✨ Improvements

• FLUX.2 Klein img2img CLI parity: Add --image-path and --image-strength to mflux-generate-flux2, enabling init-image driven generation with the same CLI pattern used in other generators.
• MLX cache control: Add --mlx-cache-limit-gb to cap MLX cache usage without requiring full --low-ram mode.

📝 Documentation

• Common CLI docs: Document --mlx-cache-limit-gb behavior and usage in the shared model README.

👩‍💻 Contributors

• @terribilissimo
• @icelaglace

✨ Improvements

• SeedVR2 directory input: Allow passing a folder to --image-path to upscale all images inside.

🧰 DX & Maintenance

• Model porting guidance: Require model README entries in the porting workflow.

📝 Documentation

• SeedVR2 usage: Document directory upscaling with CLI and Python API examples.
• CLI docs: Add Python API sections and improve Z-Image Turbo entry-point links.

✨ Improvements

• Flux2 LoRA aliasing: Add key aliases for base_model prefixes to improve LoRA resolution across configs.

📝 Documentation

• Agent guidance: Clarify skill references for Cursor agents.

🐛 Bug Fixes

• PyPI metadata: Removed invalid architecture classifier that blocked uploads (Architecture :: AArch64).

🎨 New Model Support

• Flux2 Klein (4B/9B): Full MLX port of Flux2 Klein (including multi-image editing support).
• New command: mflux-generate-flux2 for Flux2 Klein image generation.
• New command: mflux-generate-flux2-edit for Flux2 Klein image editing.

🔧 Improvements

• Qwen3-VL shared module: Extracted qwen3_vl into models/common_models/ for reuse across model families (Flux2 and Fibo etc).
• Experimental CUDA support: Added initial CUDA backend support as an experimental feature.
• Test Infrastructure: Image tests are pinned to MLX v0.30.3.

📝 Documentation

• README reorganization: Reorganized the main README for better structure and readability.

📊 Improved Metadata Handling

• Enhanced IPTC & XMP Support: Significant improvements to metadata reading and writing, ensuring better compatibility with professional image editing tools.
• Robust Metadata Extraction: Refined logic for extracting generation parameters from previously generated images.
• New Metadata Tests: Added comprehensive test suite for IPTC metadata building and original image info utilities.

🤖 DX & Maintenance

• Cursor AI Workflows: Introduced standardized Cursor commands and agent rules in .cursor/ for improved development consistency and automation.
• SeedVR2 & ControlNet Tweaks: Minor refinements to SeedVR2 and ControlNet model implementations.
• Documentation Updates: Updated README and added AGENTS.md for better contributor onboarding.

🔧 SeedVR2 Improvements

• Enhanced Color Correction: Implemented precise LAB histogram matching with wavelet reconstruction for superior color consistency between input and upscaled images.
• Configurable Softness: Added a new --softness parameter (0.0 to 1.0) to control input pre-downsampling, allowing for smoother upscaling results when desired.
• RoPE Alignment: Fixed RoPE dimension mismatch (increased to 128) to perfectly match the reference 3B transformer architecture.

🤖 DX & Maintenance

• Updated .cursorrules: Added standard procedure for test output preservation and release management.
• Updated Test Infrastructure: Updated SeedVR2 reference images and fixed dimension-related test failures.

✨ Improvements

• Better error messages for multi-file LoRA repos: When a HuggingFace LoRA repo contains multiple .safetensors files, the error message now displays copy-paste ready options instead of a raw list
• Z-Image LoRA format support: Added support for Kohya and ComfyUI LoRA naming conventions, enabling compatibility with more community LoRAs.