This release includes breaking changes for platform teams planning a safe upgrade.
✓ No known CVEs patched in this version
Topics
ReleasePort's take
Moderate signalv0.21.1 reverts breaking changes and patches critical stability issues in VOID rendering, model handling, and LTXV alignment. New features include Flux2Image, ByteDanceSeedream, OpenAI Image, and Claude LLM nodes, with cloud-runtime API endpoints now deprecated.
Why it matters: Update for stability: VOID crashes, LTXV alignment, and model handling fixes are critical. Test new Claude LLM, OpenAI Image, and Flux2Image nodes in dev. Plan migration from deprecated cloud-runtime API endpoints.
Summary
AI summaryAdd Create Video button to essentials tab, improve Hidream O1 memory usage, and fix Void runtime error.
Changes in this release
| Type | Severity | Summary | CVE |
|---|---|---|---|
| Breaking | Medium |
Revert breaking changes from previous release Revert breaking changes from previous release Source: llm_adapter@2026-05-21 Confidence: high |
— |
| Feature | Medium |
Extend Save3D node to save vertex colors and textures Extend Save3D node to save vertex colors and textures Source: llm_adapter@2026-05-21 Confidence: high |
— |
| Feature | Medium |
Support HiDream-O1-Image model Support HiDream-O1-Image model Source: llm_adapter@2026-05-21 Confidence: high |
— |
| Feature | Medium |
Add Flux2ImageNode and GrokImageEditNodeV2 partner nodes Add Flux2ImageNode and GrokImageEditNodeV2 partner nodes Source: llm_adapter@2026-05-21 Confidence: low |
— |
| Feature | Medium |
Add ByteDanceSeedreamNodeV2 partner node with Autogrow Add ByteDanceSeedreamNodeV2 partner node with Autogrow Source: llm_adapter@2026-05-21 Confidence: low |
— |
| Feature | Medium |
Add OpenAI Image partner node with DynamicCombo Add OpenAI Image partner node with DynamicCombo Source: llm_adapter@2026-05-21 Confidence: low |
— |
| Feature | Medium |
Support Anima TE LoRA Kohya format loading Support Anima TE LoRA Kohya format loading Source: llm_adapter@2026-05-21 Confidence: low |
— |
| Feature | Medium |
Add Claude LLM partner node Add Claude LLM partner node Source: llm_adapter@2026-05-21 Confidence: low |
— |
| Feature | Medium |
Add Create Video node to essentials tab Add Create Video node to essentials tab Source: llm_adapter@2026-05-21 Confidence: low |
— |
| Performance | Medium |
Improve HiDream O1 memory usage for non-dynamic VRAM Improve HiDream O1 memory usage for non-dynamic VRAM Source: llm_adapter@2026-05-21 Confidence: low |
— |
| Deprecation | Medium |
Mark cloud-runtime API endpoints as deprecated Mark cloud-runtime API endpoints as deprecated Source: llm_adapter@2026-05-21 Confidence: high |
— |
| Bugfix | Medium |
Fix VOID node RuntimeError crash Fix VOID node RuntimeError crash Source: llm_adapter@2026-05-21 Confidence: high |
— |
| Bugfix | Medium |
Fix safetensors saving of fp8 model formats Fix safetensors saving of fp8 model formats Source: llm_adapter@2026-05-21 Confidence: high |
— |
| Bugfix | Medium |
Fix data type issue with HiDream O1 model Fix data type issue with HiDream O1 model Source: llm_adapter@2026-05-21 Confidence: high |
— |
| Bugfix | Medium |
Fix LTXV mid-video multi-frame guide alignment Fix LTXV mid-video multi-frame guide alignment Source: llm_adapter@2026-05-21 Confidence: high |
— |
| Bugfix | Medium |
Create missing input directory in LoadAudio node Create missing input directory in LoadAudio node Source: llm_adapter@2026-05-21 Confidence: low |
— |
| Bugfix | Medium |
Fix Quiver partner nodes Fix Quiver partner nodes Source: llm_adapter@2026-05-21 Confidence: low |
— |
Full changelog
What's Changed
- [Partner Nodes] new Flux2ImageNode and GrokImageEditNodeV2 nodes by @bigcat88 in https://github.com/Comfy-Org/ComfyUI/pull/13814
- [Partner Nodes] new ByteDanceSeedreamNodeV2 node with DynamicCombo and Autogrow by @bigcat88 in https://github.com/Comfy-Org/ComfyUI/pull/13811
- [Partner Nodes] new OpenAI Image node with DynamicCombo and Autogrow by @bigcat88 in https://github.com/Comfy-Org/ComfyUI/pull/13838
- model_patcher: Fix safetensors saving of fp8 (CORE-160) by @rattus128 in https://github.com/Comfy-Org/ComfyUI/pull/13835
- Support anima TE lora kohya format. by @comfyanonymous in https://github.com/Comfy-Org/ComfyUI/pull/13847
- feat: Support HiDream-O1-Image (CORE-187) by @kijai in https://github.com/Comfy-Org/ComfyUI/pull/13817
- Fix dtype issue with hidream o1 by @comfyanonymous in https://github.com/Comfy-Org/ComfyUI/pull/13849
- [Partner Nodes] fix Quiver nodes by @bigcat88 in https://github.com/Comfy-Org/ComfyUI/pull/13851
- Mark deprecated cloud-runtime endpoints in spec by @MillerMedia in https://github.com/Comfy-Org/ComfyUI/pull/13789
- Suppress false-positive Spectral lint on WebSocket endpoint by @MillerMedia in https://github.com/Comfy-Org/ComfyUI/pull/13842
- Fix LTXV mid-video multi-frame guide alignment (CORE-129) by @drozbay in https://github.com/Comfy-Org/ComfyUI/pull/13625
- Revert some breaking changes. by @comfyanonymous in https://github.com/Comfy-Org/ComfyUI/pull/13861
- fix: create input directory if missing in LoadAudio define_schema by @angad777 in https://github.com/Comfy-Org/ComfyUI/pull/13834
- Better Hidream O1 mem usage factor for non dynamic vram. by @comfyanonymous in https://github.com/Comfy-Org/ComfyUI/pull/13864
- chore: update embedded docs to v0.5.0 by @comfyui-wiki in https://github.com/Comfy-Org/ComfyUI/pull/13865
- Add Create Video to the essentials tab by @AustinMroz in https://github.com/Comfy-Org/ComfyUI/pull/13863
- feat: Extend Save3D to save vertex colors and textures (CORE-189) by @kijai in https://github.com/Comfy-Org/ComfyUI/pull/13824
- [Partner Nodes] add Claude LLM node by @bigcat88 in https://github.com/Comfy-Org/ComfyUI/pull/13867
- chore: update workflow templates to v0.9.75 by @comfyui-wiki in https://github.com/Comfy-Org/ComfyUI/pull/13877
- Fix VOID failing with RuntimeError: … by @Talmaj in https://github.com/Comfy-Org/ComfyUI/pull/13873
New Contributors
- @angad777 made their first contribution in https://github.com/Comfy-Org/ComfyUI/pull/13834
Full Changelog: https://github.com/Comfy-Org/ComfyUI/compare/v0.21.0...v0.21.1
Weekly OSS security release digest.
The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.
No spam, unsubscribe anytime.
Share this release
About ComfyUI
The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface.
Related context
Related tools
Beta — feedback welcome: [email protected]