What's Changed

• Fix KSamplerAdvanced with add_noise disabled on nested latents by @kijai in #15447
• Update workflow templates to v0.11.40 by @comfyui-wiki in #15522
• chore: replace api nodes -> partner nodes in README by @robinjhuang in #15519
• Fix PreviewAny escaping non-ASCII text in dict and list previews by @christian-byrne in #15513
• Fix float64 device in ltx diffusion decoder. by @comfyanonymous in #15516
• Support anima tunes with extra blocks. by @comfyanonymous in #15555
• Don't disable dynamic vram on WSL. by @comfyanonymous in #15562
• Query pytorch for aotriton support instead of listing its lib directory by @glop102 in #15412
• Fix Generate Text ignoring thinking=false on Gemma4 E2B/E4B by @DrJKL in #15278
• Update comfy-kitchen package version to 0.2.31 by @comfyanonymous in #15564
• [Partner Nodes] feat(MiniMax): add ContextIR and Regenerate To 2K nodes by @bigcat88 in #15471
• Implement Minimax Music 3 + Core Support for Cuda Graphs by @rattus128 in #15570
• [Partner Nodes] feat(Bria): add GenFill, Eraser, Expand and Increase Resolution nodes by @bigcat88 in #15572
• llama: fix non-local x path by @rattus128 in #15580
• Update workflow templates to v0.11.41 by @comfyui-wiki in #15578
• minimax: early detect qkv vs q,k,v by @rattus128 in #15581
• Fix minimax music not working on non dynamic vram. #15588

New Contributors

• @glop102 made their first contribution in #15412

Full Changelog: v0.32.0...v0.33.1

What's Changed

• Make it easier to debug nested tensors. by @comfyanonymous in #15383
• Update workflow templates to v0.11.37 by @comfyui-wiki in #15415
• Minimum officially supported pytorch is now 2.7 by @comfyanonymous in #15413
• fix: make Create Layered Image discoverable and its flags self-explanatory by @jtydhr88 in #15429
• Fix upscale models breaking on non dynamic vram low vram. by @comfyanonymous in #15437
• Add previewable_outputs_count to /api/jobs (backend half of Media Assets badge fix) by @claude[bot] in #15148
• Extend ER-SDE noise scaler by scaling h(t) by @chaObserv in #15428
• Make a context manager for cast_bias_weight and use it. by @blepping in #14750
• Optimize MiniMax-H3 VAE by @kijai in #15446
• [Partner Nodes] feat(Qwen): add Qwen-Image 3.0 Pro by @bigcat88 in #15327
• Add tags_all / tags_any / tags_none tag filters to the assets list API by @synap5e in #15332
• Fix VAEDecodeTiled crash on NestedTensor latents (MiniMax H3) by @chelsealong in #15477
• Make cu130 warning more visible. by @comfyanonymous in #15463
• Implement comfy kitchen attention. by @comfyanonymous in #15479
• Fix peak memory issue with H3. by @comfyanonymous in #15486
• Add support for LTX 2.5 by @alexisrolland in #15499
• [Partner Nodes] feat(LTX): add new nodes for model version 2.5 by @bigcat88 in #15501
• Fix for broken tiled audio decode. by @comfyanonymous in #15502
• Update workflow templates to v0.11.39 by @comfyui-wiki in #15504
• [Partner Nodes] feat(Grok): add Grok-Imagine-Image-2.0 model by @bigcat88 in #15496
• Fix some clip vision regression. by @comfyanonymous in #15506
• Don't depend on transformers for mistral and llama tokenizers. by @comfyanonymous in #15503
• Remove potentially problematic process_tokens method. by @comfyanonymous in #15507

New Contributors

• @claude[bot] made their first contribution in #15148
• @chelsealong made their first contribution in #15477

Full Changelog: v0.31.0...v0.32.0

What's Changed

• Bump comfyui-frontend-package to 1.47.12 by @comfy-pr-bot in #15244
• Don't pin too much memory on Linux systems with no swap partition. by @comfyanonymous in #15266
• fix(minimax): cast raw parameters to input device in H3 VAEs by @rivadart in #15268
• Harden some nodes against potential issues related to combos. by @comfyanonymous in #15277
• [Partner Nodes] chore(Kling): remove retired legacy models and Virtual Try-On API by @bigcat88 in #15249
• restore SDPA non-cudnn small attention bypass by @rattus128 in #15296
• [Partner Nodes] feat(TopazAI): add Bloom 2 and Wonder 3.5 models support by @bigcat88 in #15294
• [Partner Nodes] feat(BFL): add Flux 3 video model by @bigcat88 in #15295
• Update workflow templates to v0.11.31 by @comfyui-wiki in #15297
• comfy-aimdo 0.4.12 by @rattus128 in #15290
• Update the Features section with more up to date stuff. by @comfyanonymous in #15302
• Bump comfyui-frontend-package to 1.48.6 by @comfy-pr-bot in #15301
• Make the desktop app more prominent in the readme. by @comfyanonymous in #15330
• Support int8_convrot VAE for MiniMax-H3 by @kijai in #15334
• Speedup LTX and Wan by @kijai in #15138
• Fix MiniMax H3 noise mask sampling by @endman100 in #15322
• [Partner Nodes] chore(Reve): deprecate Reve nodes by @bigcat88 in #15331
• Fix sampler issues for audio with minimax, support more samplers. by @kijai in #15243
• Skip creating useless noise in res_multistep sampler. by @comfyanonymous in #15339
• feat: Support Wan-Animate2 (CORE-358) by @kijai in #15362
• chore: fail the CodeRabbit status when the review did not run by @christian-byrne in #15382
• Fix full offload on minimax audio vae. by @comfyanonymous in #15377
• Update comfy-kitchen package version to 0.2.27 by @comfyanonymous in #15388
• [Partner Nodes] chore(Gemini,OpenAI): update pricing for image models by @bigcat88 in #15306
• Fix MiniMax H3 audio corruption with EasyCache by @kijai in #15390
• Support asym w4a8_int by @kijai in #15308
• [Partner Nodes] feat(ByteDance): add SeeDance 2.5 model by @bigcat88 in #15395
• Bump comfyui-frontend-package to 1.48.7 by @comfy-pr-bot in #15403
• [Partner Nodes] feat: ImageCompositor node with layer-state compositing, layer from bbox and Seedream Layer Separation node by @jtydhr88 in #15317
• Update workflow templates to v0.11.34 by @comfyui-wiki in #15404
• Update comfy-kitchen package version to 0.2.28 by @comfyanonymous in #15407

New Contributors

• @rivadart made their first contribution in #15268
• @endman100 made their first contribution in #15322

Full Changelog: v0.30.0...v0.31.0

What's Changed

• Update comfy-kitchen package version to 0.2.23 by @comfyanonymous in #15112
• Support int8 convrot embedding lookup (CORE-371) by @kijai in #15035
• Load weights to process RAM with MRU policy using pinning infrastructure by @rattus128 in #15027
• Add configurable DETAIL logging side channel (CORE-369) by @rattus128 in #15064
• Support PrunaVAED (faster LTX2.3 decoder) by @kijai in #15129
• [Dataset/Security,Feature] Add dataset folder to avoid arbitrary folder access for dataset stuff. by @KohakuBlueleaf in #14807
• Fix LTXAV crash when sampling without an audio latent by @drozbay in #15132
• Update comfy-kitchen to fix flux kv issue. by @comfyanonymous in #15144
• Fallback to cudnn attention on linux if flash attention doesn't work. by @comfyanonymous in #15146
• Fix user.css loading broken by #14734 by @kaalibro in #15000
• fix(jobs): prefer media over text for job preview_output by @DeniDoman in #14681
• Fix SVG previews broken by the stored-XSS forced-download by @mattmillerai in #15149
• Don't enable comfyui_detail.log by default. by @comfyanonymous in #15159
• comfy-kitchen AMD support. by @comfyanonymous in #15160
• Bump comfyui-frontend-package to 1.47.11 by @comfy-pr-bot in #15161
• [Partner Nodes] adjust pricing for the Bria Video RemoveBackground nodes by @bigcat88 in #15155
• [Partner Nodes] feat(Ideogram): add new P-Image model support by @bigcat88 in #15154
• Update workflow templates to v0.11.20 by @comfyui-wiki in #15165
• [Partner Nodes] feat(Minimax): add new MiniMax H3 model support by @comfyui-wiki in #15167
• fix: resend cached histogram UI for CurveEditor after page refresh by @jtydhr88 in #15152
• Don't assume sampler_function has a name attribute in detail logging by @blepping in #15179
• [Partner Nodes] chore(Bria): adjust pricing for other video endpoints by @bigcat88 in #15171
• Delay dynamic pin cleanup until model destruction by @rattus128 in #15183
• ops: apply the custom placeholder logic to Linux too by @rattus128 in #15181
• Update workflow templates to v0.11.23 by @comfyui-wiki in #15187
• [Partner Nodes] fix(ByteDance): encode stereo reference audio without doubling its duration by @bigcat88 in #15177
• Expand k, v when attention backend would fall back to math because gqa. by @comfyanonymous in #15190
• Add crf option to save video node. by @comfyanonymous in #15191
• Store mp4 metadata at the beginning of the file when possible. by @comfyanonymous in #15195
• [Partner Nodes] feat(xAI): update nodes for grok-imagine-video-1.5 model by @bigcat88 in #15197
• [Partner Nodes] chore(Bria): increase price for video endpoints by @bigcat88 in #15186
• Update comfy-kitchen version to 0.2.26 by @comfyanonymous in #15208
• Let the VAEDecodeAudio node decode nested audio. by @comfyanonymous in #15211
• Support latent previews for nested latents by @kijai in #15196
• Use the actual function to check if the path is within the dir. by @comfyanonymous in #15216
• Let the save text node save in csv format. by @comfyanonymous in #15217
• Disable gradients on diffusion models. by @comfyanonymous in #15218
• Handle case where swap memory query fails on windows. by @comfyanonymous in #15219
• comfy-aimdo 0.4.11 by @rattus128 in #15215
• [Partner Nodes] feat(Minimax): add 768P resolution for H3 model by @bigcat88 in #15227
• fix pin registration priority by @rattus128 in #15226
• rename comfy/logging.py to comfy/internal_logging.py by @olifre in #15231
• feat: Support MiniMax-H3 (CORE-375) by @kijai in #15224
• Update workflow templates to v0.11.26 by @comfyui-wiki in #15228
• Update workflow templates to v0.11.27 by @comfyui-wiki in #15239

New Contributors

• @kaalibro made their first contribution in #15000
• @DeniDoman made their first contribution in #14681
• @olifre made their first contribution in #15231

Full Changelog: v0.29.0...v0.30.0

Frontend fixes and new api/partner nodes.

Full Changelog: v0.29.0...v0.29.2

What's Changed

• fix(Video): stream the video transcode instead of buffering every frame in RAM (CORE-353) (CORE-351) by @bigcat88 in #14813
• [Partner Nodes] feat(client): send ComfyUI Job Id in request headers by @alexisrolland in #14934
• Update AGENTS.md by @comfyanonymous in #14955
• [feat]Add JoyImageEdit native model support by @feice-huang in #14428
• [Partner Nodes] feat(OpenAI): add GPT5.6 models by @bigcat88 in #14957
• Various comfy kitchen optimizations and fixes. by @comfyanonymous in #14963
• Speed up anima a bit. by @comfyanonymous in #14953
• Support anima lllite control models. by @comfyanonymous in #14954
• [Partner Nodes] feat(Google): add Gemini 3.5 Flash LLM model by @bigcat88 in #14972
• chore(openapi): sync shared API contract from cloud@e964994 by @comfy-pr-bot in #14947
• Update workflow templates to v0.11.11 by @comfyui-wiki in #14973
• [Partner Nodes] feat(HeyGen): add Avatar, Talking Photo, Create Avatar, Video Translate and TTS nodes by @bigcat88 in #14958
• More comfy-kitchen int8 optimizations. by @comfyanonymous in #14980
• Update workflow templates to v0.11.12 by @comfyui-wiki in #14986
• [Partner Nodes] fix(Google): switch to the Interactions API for Omni model by @bigcat88 in #14984
• Implement regular and timestep zero reference images to krea 2 for ostris and identity edit ref loras. by @comfyanonymous in #14843
• Fix wan dancer issue with batches. by @comfyanonymous in #14999
• feat: Support Gemma4 12B (CORE-277) by @kijai in #14304
• Fix gfx1035 not being treated like RDNA2 by @comfyanonymous in #15009
• chore: add /AGENTS.md to CODEOWNERS by @mattmillerai in #14962
• [Trainer,Dataset/Feature] Video processing nodes, Image Processing Node video support, trainer video support (CORE-81) by @KohakuBlueleaf in #13588
• FreSca: 5D+ (ex. Anima) fix, model-agnostic iteration by @TheToxin-git in #15007
• Add native Uni3C Controlnet support for Wan models (CORE-365) by @drozbay in #14946
• [Partner Nodes] fix(Gemini-Omni): pass videos as inline data by @bigcat88 in #15014
• ci: bump cursor-review to github-workflows@964d5aa by @cloud-code-bot[bot] in #15017
• Use optimized rms_rope function in joyai image model. by @comfyanonymous in #15018
• [Partner Nodes] feat(Openrouter): add new models by @bigcat88 in #15021
• [Partner Nodes] feat(Anthropic): add new models by @bigcat88 in #15023
• Update workflow templates to v0.11.15 by @comfyui-wiki in #15030
• Make z image/lumina 2 models use comfy kitchen rms rope. by @comfyanonymous in #15036
• Bump comfyui-frontend-package to 1.47.10 by @comfy-pr-bot in #15045
• [Partner Nodes] feat(ByteDance): add new audio-1.0-multilingual model by @bigcat88 in #15034
• [Partner Nodes] chore(Runway): deprecate Gen3a model by @bigcat88 in #15050
• Make Ernie use comfy kitchen rms rope by @comfyanonymous in #15055
• Update workflow templates to v0.11.17 by @comfyui-wiki in #15059
• Use comfy kitchen rope functions in ltx models. by @comfyanonymous in #15056
• upscalers: convert latent_upsampler model to ModelPatcherDynamic by @rattus128 in #15063
• cli_args: bump default background usage clamp to 128GB by @rattus128 in #15068
• feat: Support MageFlow (CORE-372) by @kijai in #15026
• Improve LTXV IC-lora detection by @kijai in #15073
• [Partner Nodes] feat(OpenRouter): add Claude Opus 5 by @bigcat88 in #15075
• Optimize ideogram model using comfy kitchen rms rope. by @comfyanonymous in #15080
• Fix MageFlow on cards that don't support bf16 by @kijai in #15081
• Update AGENTS.md by @comfyanonymous in #15096
• Update embedded docs to v0.5.9 by @comfyui-wiki in #15092
• [Partner Nodes] feat(Anthropic): add Claude Opus 5 model by @bigcat88 in #15079
• Allow using float fps for LTXVEmptyLatentAudio by @kijai in #15106
• Update stable portable release workflow. by @comfyanonymous in #15113
• [Partner Nodes] feat(credits): respect "X-Comfy-Credits-Used" header from the comfy-api by @bigcat88 in #15091
• [Partner Nodes] feat(Recraft): add V4.1 models and V4 image edit nodes by @bigcat88 in #15105
• Update workflow templates to v0.11.19 by @comfyui-wiki in #15123
• Go back to older rocm for portable. by @comfyanonymous in #15127

New Contributors

• @feice-huang made their first contribution in #14428
• @TheToxin-git made their first contribution in #15007
• @cloud-code-bot[bot] made their first contribution in #15017

Full Changelog: v0.28.0...v0.29.0

What's Changed

• Add AGENTS.md by @comfyanonymous in #14696
• Add some more stuff to AGENTS.md by @comfyanonymous in #14704
• Fix Qwen3-VL tokenizer crash with custom embeddings by @silveroxides in #14713
• Update AGENTS.md with more stuff. by @comfyanonymous in #14725
• [Partner Nodes] chore(Ideogram): remove IdeogramV1 and IdeogramV2 nodes by @bigcat88 in #14712
• Update AGENTS.md by @comfyanonymous in #14726
• security: fix four vulnerabilities (GHSA-779p-m5rp-r4h4) by @mattmillerai in #14734
• Update AGENTS.md by @comfyanonymous in #14733
• [Partner Nodes] feat(ByteDance): add support for Seed Audio 1.0 by @bigcat88 in #14731
• Update workflow templates to v0.11.2 by @comfyui-wiki in #14741
• [Partner Nodes] chore(StabilityAI): remove StabilityAI nodes by @bigcat88 in #14737
• Update AGENTS.md by @comfyanonymous in #14738
• [Partner Nodes] chore(ByteDance): adjust category name by @bigcat88 in #14752
• Add CLAUDE.md as symlink to AGENTS.md by @robinjhuang in #14757
• fix: error when embedding is loaded with models using llama_template by @silveroxides in #14744
• Try to make coderabbit enforce AGENTS.md by @comfyanonymous in #14759
• fix: Bug when setting transparency in color picker by @alexisrolland in #14764
• [Partner Nodes] fix(logs-auth): mask authorization headers in logs by @bigcat88 in #14774
• Small speedup for text model sampling. by @comfyanonymous in #14773
• Update embedded docs to v0.5.7 by @comfyui-wiki in #14783
• Skip unloading dynamic model patchers in current workflow. by @comfyanonymous in #14799
• ci: add CLA Assistant workflow by @alexisrolland in #14582
• Delete symlink that breaks our updates. by @comfyanonymous in #14803
• fix(Video): don't crash on videos with undecodable audio streams by @bigcat88 in #14746
• feat: Implement basic text overlay node (CORE-137) by @drozbay in #14610
• Update AGENTS.md by @comfyanonymous in #14819
• Support gqa on all attention backends, drop support for pytorch 2.4 by @comfyanonymous in #14772
• Fix crash when using UNetSelfAttentionMultiply by @comfyanonymous in #14823
• add models-directory launch argument by @silveroxides in #9113
• Fix AttributeError in VAE.is_dynamic() for VAEs constructed without a patcher by @j2gg0s in #14826
• [Partner Nodes] feat(ByteDance): add Seedream 5 Pro model support by @bigcat88 in #14832
• Update workflow templates to v0.11.6 by @comfyui-wiki in #14834
• feat(assets): add namespaced model_type tags and align tag semantics by @synap5e in #14511
• Fix qwen3vl reference images when used as a text encode models. by @comfyanonymous in #14845
• feat: allow --comfy-api-base target ephemeral testenvs by @bigcat88 in #14569
• fix: Load3D failing path validation from double path resolution by @jtydhr88 in #14852
• feat(api): expose effective extension filter on /experiment/models by @synap5e in #14797
• Update cla.yml by @alexisrolland in #14851
• Support convrot int4 models. by @comfyanonymous in #14859
• Try to fix the model reloading issue some people have. by @comfyanonymous in #14822
• Fix FP8 activation quantization for >2D activations in mixed_precision_ops by @liminfei-amd in #14643
• Enable comfy-kitchen Triton backend by default on ROCm/AMD by @liminfei-amd in #14862
• Fix black image on turing when using int4 models. by @comfyanonymous in #14864
• Add SeedVR2 support (CORE-6) by @pollockjj in #14424
• [Partner Nodes] feat(Seedream): add widget to disable thinking by @bigcat88 in #14853
• [Partner Nodes] fix(Tencent): restore Tencent3DPartNode FBX output by @bigcat88 in #14867
• Only auto-enable the ROCm comfy-kitchen Triton backend on matrix-core GPUs by @liminfei-amd in #14869
• Save Text Node (CORE-176) by @yousef-rafat in #14102
• feat: add bboxes input to Create Bounding Boxes node by @jtydhr88 in #14724
• CORE-329 feat: add Save 3D (Advanced) node family by @jtydhr88 in #14701
• fix: Undeprecate the text node by @alexisrolland in #14870
• chore: Update preview nodes by @alexisrolland in #14871
• Try to fix some issues with the seedvr VAE. by @comfyanonymous in #14877
• Temporarily disable auto enabling triton by default on AMD. by @comfyanonymous in #14878
• Try to fix flash attention related issue on AMD. by @comfyanonymous in #14880
• Fix SaveVideo description: says images, saves video by @gmhelmold in #14885
• Support PID 1.5 models. by @comfyanonymous in #14894
• [Partner Nodes] feat(client): send ComfyUI Core version in request headers by @bigcat88 in #14910
• fix(image): correct HLG inverse-OETF clamp in hlg_to_linear by @bigcat88 in #14762
• fix(image): support single-channel images in Save Image (Advanced) by @bigcat88 in #14761
• [Partner Nodes] chore(Google): reroute Gemini Image preview models to release versions by @bigcat88 in #14917
• Update embedded docs to v0.5.8 by @comfyui-wiki in #14920
• chore(openapi): sync shared API contract from cloud@491db61 by @comfy-pr-bot in #14815
• Update workflow templates to v0.11.9 by @comfyui-wiki in #14924
• Fix hidream o1 regression. by @comfyanonymous in #14923
• More optimized int8 and int4 on turing. by @comfyanonymous in #14927
• Fix 3D advanced nodes crashing in API mode (no UI) by @jtydhr88 in #14930
• Fix int8 performance regression on 16xx series. by @comfyanonymous in #14941
• [Partner Nodes] feat(sync.so): add support for "sync-3" model by @bigcat88 in #14928
• Fix cached outputs missing from job results when prompt has no client_id by @bigcat88 in #14939
• Bump comfyui-frontend-package to 1.45.21 by @comfy-pr-bot in #14944

New Contributors

• @j2gg0s made their first contribution in #14826
• @liminfei-amd made their first contribution in #14643
• @gmhelmold made their first contribution in #14885

Full Changelog: v0.27.0...v0.28.0

The major change in this version is added support for int8 convrot models.

What's Changed

• chore(openapi): sync shared API contract from cloud@363764b by @comfy-pr-bot in #14607
• [Partner Nodes] feat(Alibaba): add support for HappyHorse 1.1 model by @bigcat88 in #14611
• Downgrade the template to 0.10.2 to bring the HH1.1 templates back by @comfyui-wiki in #14613
• [Partner Nodes] feat(Grok): add 1080p resolution to Grok Image node by @bigcat88 in #14612
• [Partner Nodes] feat(ByteDance): add 4K resolution support for SeeDance 2.0 by @bigcat88 in #14614
• feat: Add support for Seed node (CORE-295) by @yousef-rafat in #14616
• Add advanced krea 2 model merging node. by @comfyanonymous in #14621
• chore(openapi): sync shared API contract from cloud@42eb0d1 by @comfy-pr-bot in #14619
• feat: Bounding box canvas and Ideogram JSON prompt (CORE-292) by @jtydhr88 in #14537
• [Partner Nodes] feat(ByteDance): add support for SeeDance-2.0-Mini video model by @bigcat88 in #14626
• Update workflow templates to v0.10.7 by @comfyui-wiki in #14632
• Support int8 models. by @comfyanonymous in #14636
• Faster int8. by @comfyanonymous in #14641
• Fix int8 loras causing lower quality requant with wrong settings. by @comfyanonymous in #14650
• Update GLSL node to use ANGLE library (CORE-162) by @pythongosssss in #13195
• int8 support on turing GPUs. by @comfyanonymous in #14662
• ci: add team-gated Cursor review (thin caller for github-workflows) by @mattmillerai in #14527
• Fix int8 mm being skipped on offloaded lora weights. by @comfyanonymous in #14669
• Update embedded docs to v0.5.6 by @comfyui-wiki in #14668
• Better and faster int8 lora applying. by @comfyanonymous in #14685
• Bump comfyui-frontend-package to 1.45.20 by @comfy-pr-bot in #14684
• Add ConditioningMultiply node to nodes.py as an addition to other adj… by @silveroxides in #14686
• [Partner Nodes] feat(Google): add Nano Banana 2 Lite model by @bigcat88 in #14693
• Fix memory leak related to int8. by @comfyanonymous in #14697
• feat: make asset hashing opt-in via --enable-asset-hashing (off by default) by @mattmillerai in #14663
• [Partner Nodes] feat(Google): add Gemini Video Omni node by @bigcat88 in #14695
• chore: Update nodes categories by @alexisrolland in #14674
• Update workflow templates to v0.11.1 by @comfyui-wiki in #14698

Full Changelog: v0.26.0...v0.27.0

What's Changed

• chore(openapi): sync shared API contract from cloud@00ef9cc by @comfy-pr-bot in #14423
• [Partner Nodes] chore(SoniloTextToMusic): reduce price by half by @bigcat88 in #14500
• Add the checkbox to disable head drawing in node SDPoseDrawKeypoints by @rmarma in #14446
• fix: log base directory to startup messages when --base-directory is used by @octo-patch in #13370
• feat: Support text generation with Qwen3-VL (CORE-276) by @kijai in #14298
• chore: Update nodes categories (CORE-263) by @alexisrolland in #14460
• Add new model blueprints by @comfyui-wiki in #14506
• docs: add M3 and M4 to Apple Silicon supported chips list by @EXA4V in #14449
• Update link to new comfyui desktop repo. by @comfyanonymous in #14516
• Update TripoSplat nodes categories by @alexisrolland in #14512
• feat: SCAIL-2 multireference (CORE-310) by @kijai in #14509
• Allow using Qwen3-VL as flux2 klein text encoder (again) by @kijai in #14526
• feat: Support Boogu-Image (CORE-308) by @kijai in #14523
• Add negative prompt to boogu edit node and set min images to 0. by @comfyanonymous in #14529
• [Partner Nodes] feat(Kling): add support for Kling V3-Turbo model by @bigcat88 in #14528
• feat: add enable_telemetry CLI feature flag by @Kosinkadink in #14530
• Update ComfyUI Desktop to Comfy Desktop for consistent product naming by @comfyui-wiki in #14533
• chore(openapi): sync shared API contract from cloud@d10ff72 by @comfy-pr-bot in #14518
• harden: load training-dataset shards with weights_only=True by @mattmillerai in #14543
• Move comfy sys path insert to custom node loading. by @comfyanonymous in #14459
• [Partner Nodes] feat(Luma): add support for Luma Rays 3.2 by @bigcat88 in #14540
• [Partner Nodes] fix: respect Retry-After header by @bigcat88 in #14234
• Bump comfyui-frontend-package to 1.45.19 by @comfy-pr-bot in #14559
• feat: add Load3DAdvanced node by @jtydhr88 in #14316
• Add jobs-namespace cancel endpoints (POST /api/jobs/{job_id}/cancel, POST /api/jobs/cancel) by @mattmillerai in #14493
• feat: Context Windows sampling with LTX2 models and IC-LoRa guides (CORE-3) by @drozbay in #13325
• Rename a bunch of nodes by @alexisrolland in #14547
• Small anima optimization. by @comfyanonymous in #14557
• chore(openapi): sync shared API contract from cloud@1aea581 by @comfy-pr-bot in #14562
• [Partner Nodes] chore(Google): remove preview versions of models that will be deprecated soon by @bigcat88 in #14555
• Update extra model paths example. by @comfyanonymous in #14570
• Add output socket to save nodes by @alexisrolland in #13866
• Update embedded docs to v0.5.5 by @comfyui-wiki in #14585
• feat: Support Krea2 by @kijai in #14589
• Please try native formats instead of disabling dynamic vram. by @comfyanonymous in #14577
• More accurate memory usage factor for krea 2. by @comfyanonymous in #14594
• [Partner Nodes] feat(Alibaba): add support for HappyHorse 1.1 model by @bigcat88 in #14581
• [Partner Nodes] feat(ByteDance): add 4K resolution support for SeeDance 2.0 by @bigcat88 in #14588
• Update workflow templates to v0.10.2 by @comfyui-wiki in #14600
• [Partner Nodes] feat(Grok): add 1080p resolution to Grok Image node by @bigcat88 in #14597
• [Partner Nodes] revert last 3 PRs by @bigcat88 in #14602
• Update workflow templates to v0.10.3 by @comfyui-wiki in #14603

New Contributors

• @rmarma made their first contribution in #14446
• @EXA4V made their first contribution in #14449

Full Changelog: v0.25.0...v0.26.0

• [Partner Nodes] feat(Kling): add support for Kling V3-Turbo model (#14528).

Full Changelog: v0.25.0...v0.25.1