v0.35.0

What's Changed

• [Partner Nodes] chore(Google): drop retiring Veo 2 and Veo 3.0 models by @bigcat88 in #15883
• [Partner Nodes] feat(Recraft): add V4 Styles by @bigcat88 in #15903
• [Partner Nodes] feat(WAN): add WAN3-Prime model support by @bigcat88 in #15894
• Bump comfyui-frontend-package to 1.51.9 by @comfy-pr-bot in #15696
• Support avif in Save Image Advanced node. by @comfyanonymous in #15891
• fix(video): remux HEVC to mp4/mov as hvc1 via hevc_mp4toannexb by @bigcat88 in #15809
• fix(memory): respect the container cgroup memory limit instead of host RAM (CORE-394) by @bigcat88 in #15927
• Update workflow templates to v0.11.50 by @comfyui-wiki in #15931
• [Partner Nodes] feat(Google-Omni): add support for Omni 1.1 model by @bigcat88 in #15929
• feat(3d): File3DToMesh node — parse GLB/GLTF/OBJ/STL into MESH by @jtydhr88 in #15919
• Add to readme that we support HDR and high bit depth. by @comfyanonymous in #15940
• Add cfgpp_ud10_ab sampler. by @comfyanonymous in #15951
• MiniMax-H3: Support PDD LoRA by @kijai in #15908
• Add section about user input tolerance to AGENTS.md by @comfyanonymous in #15955
• [Partner Nodes] fix(HeyGen): update Avatar Video price badge by @bigcat88 in #15945
• Improve some warning messages. by @comfyanonymous in #15977
• feat: add VideoTrim and VideoCrop nodes with VIDEO_EDIT widget inputs by @jtydhr88 in #15637
• Disable push event trigger for failing workflow. by @comfyanonymous in #15995
• Add how to cache model stuff to AGENTS.md by @comfyanonymous in #16003
• Minimax h3 controlnet as a model patch instead of a controlnet. by @comfyanonymous in #15975
• Support SenseNova U1.5 (CORE-411) by @T8mars in #15922
• Allow using references with MiniMax-H3 Fun Union and fix prefetch rac… by @kijai in #16020
• Disable ck attention when fp32 + low_precision_attention False. by @comfyanonymous in #16029
• Enable PyTorch SDPA attention and FP8 ops for gfx1170 by @tvukovic-amd in #16014
• [Partner Nodes] feat(MiniMax): add H3 Max option to H3 text-to-video and first-last-frame nodes by @purzbeats in #16025
• Update workflow templates to v0.11.52 by @comfyui-wiki in #16024
• Lower trellis workflow remesh memory usage. by @comfyanonymous in #16034
• Annotate temp paths in Preview3DAdvanced / PreviewGaussianSplat / PreviewPointCloud outputs by @jtydhr88 in #16038
• [Partner Nodes] feat(Minimax-H3): add the Max model to the Reference node by @bigcat88 in #16041
• Update workflow templates to v0.11.54 by @comfyui-wiki in #16045
• Fix SAM3 segmentation head receiving 4 FPN levels when scalp=1 by @ussoewwin in #15979
• Enable PyTorch SDPA attention and FP8 ops for gfx1171 by @tvukovic-amd in #16039
• Add live resolution preview widget to ResolutionSelector by @jtydhr88 in #16013
• Emit the document canvas from ImageCompositor for the layer editor by @jtydhr88 in #15742
• Add warning about forcing single GPU mode on Windows by @comfyanonymous in #16055
• Register partner_run_gate_enabled feature flag by @MaanilVerma in #16012
• Update README.md by @alexisrolland in #16071
• MiniMax-H3: Make VAE optional, allow using text encoder only references by @kijai in #16065
• Support MiniMax-H3 and HiDream O1 loras trained with DiffSynth-Studio / ModelScope by @yjy415 in #15662
• Update nodes to improve search rank by @alexisrolland in #16074
• Lower peak trellis workflow vram and ram usage. by @comfyanonymous in #16054
• Make comfyui handle missing workflow media properly. by @comfyanonymous in #16079
• Remove some workflow media from portable install to make it smaller. by @comfyanonymous in #16080
• [Partner Nodes] feat(Recraft): add V4 Styles Pro models to Create Style node by @bigcat88 in #16043
• Fix Get Image Size Description by @rattus128 in #16091
• Allow disconnected Switch inputs by @rattus128 in #16083
• chore: Update nodes categories by @alexisrolland in #16077
• Introduce Comfy Compiler (CORE-389) by @rattus128 in #15861
• Remove now unecessary minimax memory workaround. by @comfyanonymous in #16103
• [Partner Nodes] fix(Tripo): real output formats, exact pricing, missing parameters and six new nodes by @bigcat88 in #16060
• [Partner Nodes] feat(Meta): add Muse Image nodes by @purzbeats in #16078
• [Partner Nodes] feat(Anthropic): add Claude Fable 5.1 to the Claude node by @bigcat88 in #16085
• [Partner Nodes] feat(MiniMax): add H3 Max Turbo option to H3 text-to-video and first-last-frame nodes by @bigcat88 in #16094
• Update workflow templates to v0.11.55 by @comfyui-wiki in #16104
• comfy-aimdo 0.5.2 by @rattus128 in #16108
• [Partner Nodes] feat(Comfy-Cloud): add nodes that run curated workflows on Comfy Cloud GPUs by @deepme987 in #15935
• [Partner Nodes] feat(OpenAI): add gpt 6 astra model and reasoning param by @bigcat88 in #16125
• [Partner Nodes] deprecate retired models by @bigcat88 in #16121
• Update comfy-kitchen version to 0.2.33 by @comfyanonymous in #16133
• Fix potential mps issue of supports_fp64 called with None. by @comfyanonymous in #16136
• Add Sparse Attention node by @kijai in #16072
• Update embedded docs to v0.5.11 by @comfyui-wiki in #15947
• Pause comfy compiler for long lived sparse allocations by @rattus128 in #16148
• fix(tests): Stop folder_path_test from replacing the shared cli_args singleton by @guill in #16053
• Fix the RGB/YUV conversion nodes on images with an alpha channel by @christian-byrne in #15632
• Add color space conversion node. by @comfyanonymous in #16135
• Fix Detect Edges (Canny) on images with an alpha channel by @christian-byrne in #15631
• Don't adjust the alpha channel in the image color adjustment nodes by @christian-byrne in #15629
• Fix Quantize Image on images with an alpha channel by @christian-byrne in #15630
• Fix Draw Text Overlay on images with an alpha channel by @christian-byrne in #15634
• Don't blend the alpha channel in the Blend Images node by @christian-byrne in #15625
• Don't invert the alpha channel in the Invert Image node by @christian-byrne in #15622
• Fix Porter-Duff blend modes with transparent layers by @tandede in #15721
• Bump comfyui-frontend-package to 1.51.10 by @comfy-pr-bot in #16118
• Disable int8 weight-only quantization on devices without torch._int_mm (fixes MPS crash) by @chelsealong in #16130
• Fix registration issues by @guill in #15890
• Fix: Verify aotriton kernels actually launch before enabling pytorch attention by @0xDELUXA in #15648
• Remove useless code. by @comfyanonymous in #16169
• chore: Harmonize model attention nodes by @alexisrolland in #16154
• Add LTXV generated-keyframe nodes and Freeze Latent by @adamoster in #16040
• Comfy Aimdo 0.5.3 + Memory compiler fixes by @rattus128 in #16180
• feat: Pixal3D multiview model support (CORE-421) by @kijai in #16048
• Fix quantized text encoder matmul gating and Gemma4 prefill cache by @kijai in #16185
• Cleaner way of enabling quantized mm on text gen but not on text enc. by @comfyanonymous in #16189
• Add LTXVAddLatentGuide for pinning a pre-encoded latent as a guide by @alexanderar in #16176
• Fix MiniMax H3 denoise mask velocity conversion by @poorpaper in #15988
• Only lock repo PRs after merging if they contain a CLA signature. by @comfyanonymous in #16191
• [Partner Nodes] feat(OpenAI): add GPT Image 2.5 Flare and Sunburst models by @bigcat88 in #16190
• chores: Update tooltips of 3D nodes (CORE-323) by @alexisrolland in #16179
• Update workflow templates to v0.11.57 by @comfyui-wiki in #16192
• Report the file saved by Save 3D (Advanced) as a standard 3d output item by @jtydhr88 in #16171

New Contributors

• @T8mars made their first contribution in #15922
• @tvukovic-amd made their first contribution in #16014
• @ussoewwin made their first contribution in #15979
• @MaanilVerma made their first contribution in #16012
• @yjy415 made their first contribution in #15662
• @tandede made their first contribution in #15721
• @adamoster made their first contribution in #16040
• @alexanderar made their first contribution in #16176
• @poorpaper made their first contribution in #15988

Full Changelog: v0.34.0...v0.35.0