Add/update changelog/index.mdx

Author: coderfromthenorth93

changelog/index.mdx

@@ -4,6 +4,49 @@ description: "Track ComfyUI's latest features, improvements, and bug fixes. For
 icon: "clock-rotate-left"
 ---
 
+<Update label="v0.5.0" description="December 17, 2025">
+
+**Frontend & API Updates**
+- Frontend updated to v1.34.9 with enhanced workflow handling and improved file endpoint security
+- Per-queue live preview method override allows for customized preview experiences across different workflows
+- Workflow templates updated to v0.7.59 for better compatibility with modern node structures
+
+**Memory & Performance Optimizations**
+- Improved memory estimation for Z-Image and SD3 models, reducing VRAM allocation issues
+- Enhanced VAE loading requirements for better GPU memory management, particularly benefiting Qwen-Image VAE workflows
+- Fixed dtype handling in mixed operations, preventing bias-related errors during model execution
+- Optimized chroma radiance VRAM estimation for more efficient rendering workflows
+
+**Model Support Enhancements**
+- WanMove model support added for advanced motion processing workflows
+- New Qwen Edit 2511 reference method support with automatic detection based on state dict keys
+- Z-Image Fun Control Union 2.0 basic implementation with inpainting capabilities through ZImageFunControlnet node
+- Improved Z-Image compatibility with FP16 on newer PyTorch versions
+
+**API Node Updates**
+- Kling Omni Image node now available for image generation workflows
+- Added Kling TextToVideoWithAudio and ImageToVideoWithAudio nodes for comprehensive video generation
+- Wan2.6 model integration for enhanced video processing capabilities
+- Deprecated older Pika API nodes and legacy Kling models for streamlined experience
+
+**Sampling & Generation**
+- New SEEDS_2 phi_2 variant and dedicated sampler node for enhanced generation control
+- Added exp_heun_2_x0 sampler series providing additional sampling options for fine-tuned outputs
+- Context windows callback support for custom conditioning handling in advanced workflows
+
+**3D & External Integrations**
+- Tripo3.0 support added for 3D model generation workflows
+- Enhanced Lumina and Wan model block information for better workflow debugging and optimization
+
+**Bug Fixes & Stability**
+- Fixed Node V1 dictionary result processing in V3 node implementations
+- Resolved portable updater issues with unmerged repository states
+- Corrected hidden file filtering from internal file endpoints
+- Fixed PyTorch warnings and deprecated type usage for cleaner operation
+- Improved torch.compile compatibility by disabling guards on transformer_options when needed
+
+</Update>
+
 <Update label="v0.4.0" description="December 10, 2025">
 
 **Model Support**
@@ -93,7 +136,6 @@ icon: "clock-rotate-left"
 - Enhanced Flux2 reference image memory estimation to reduce VRAM errors
 
 </Update>
-
 <Update label="v0.3.73" description="November 26, 2025">
 
 **Bug Fixes**