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

What's Changed

• chore(openapi): sync shared API contract from cloud@7c470f0 by @comfy-pr-bot in #14174
• fix: Image grid bug fix (CORE-215) by @yousef-rafat in #14100
• [Partner Nodes] allow SaveWEBM node to save with alpha channel; add BriaTransparentVideoBackground node by @bigcat88 in #14257
• [Partner Nodes] fix (Seedance 2.0): prevent 1080p first/last-frame stretch jump by @bigcat88 in #14251
• [Partner Nodes] feat: add Bria Green Background node by @bigcat88 in #14277
• [Partner Nodes] feat: add Krea 2 Medium Turbo model by @bigcat88 in #14280
• [Partner Nodes] feat: add seed input to Flux Erase node by @bigcat88 in #14283
• Update workflow templates to v0.9.98 by @comfyui-wiki in #14284
• Bump comfyui-frontend-package to 1.45.15 by @comfy-pr-bot in #14265
• Fix ideogram if model dtype gets set to fp8. by @comfyanonymous in #14291
• Consolidate audio nodes into SaveAudioAdvanced node (CORE-202) by @alexisrolland in #13871
• Enable cfg1 optimization for DualModelGuider with CFGGuider by @kijai in #14290
• Fix interoperation with external source of pinned memory pressure by @rattus128 in #14252
• Comfy Aimdo 0.4.9 by @rattus128 in #14300
• [Partner Nodes] feat: add new Gemini text node by @bigcat88 in #14299
• [Partner Nodes] feat: add temperature and top_p to NanoBanan node by @bigcat88 in #14305
• feat: add PreviewGaussianSplat + PreviewPointCloud nodes by @jtydhr88 in #14194
• Update AMD portable readme. by @comfyanonymous in #14303
• BE-1172 fix(3d): let Load3D accept output/temp paths, save Preview3DAdvanced to temp/ by @jtydhr88 in #14294
• feat(3d): reorder Preview3DAdvanced / PreviewGaussianSplat / PreviewPointCloud inputs and outputs by @jtydhr88 in #14308
• Update line endings check to ignore .ci files. by @comfyanonymous in #14319
• Use windows line endings for windows portable readmes. by @comfyanonymous in #14334
• Add SeedVR2 support (CORE-6) by @pollockjj in #14110
• Update embedded docs to v0.5.3 by @comfyui-wiki in #14350
• Add Color primitive by @kijai in #14260
• Improve ResolutionSelector by @silveroxides in #14309
• feat(assets): extract image dimensions at ingest and emit on asset responses by @mattmillerai in #13991
• Revert "Add SeedVR2 support (CORE-6)" by @comfyanonymous in #14359
• chore(openapi): sync shared API contract from cloud@5273c30 by @comfy-pr-bot in #14266
• fix: Add back apply_rotary_emb for Qwen Image by @alexisrolland in #14364
• Allow custom templates with Ideogram4 TE by @kijai in #14374
• main/server: Add --debug-hang by @rattus128 in #14371
• Add LoRA key mapping for LTXV/LTXAV models by @kelseyee in #14349
• feat: Add model support for SCAIL-2 (CORE-207) by @kijai in #14373
• chore: Move bg_removal_model input socket to first position for nicer display by @alexisrolland in #14353
• Fix ideogram crash: mm: dont reset cast buffers in cleanup_models_gc() by @rattus128 in #14372
• [Trainer/bug] Avoid conditions become trainable in training node by @KohakuBlueleaf in #14368
• feat: Add Bernini-R model support (Wan video) (CORE-279) by @kijai in #14216
• Depth anything 3 (Core-135) by @Talmaj in #13853
• Always enable cuda malloc on cu130 and higher. by @comfyanonymous in #14381
• chore(openapi): sync shared API contract from cloud@ca12913 by @comfy-pr-bot in #14367
• [Trainer/bug] Ensure model is not inference mode (CORE-72) by @KohakuBlueleaf in #13400
• chore(assets): drop vestigial tags.tag_type column by @mattmillerai in #14248
• feat(assets): cursor-based pagination on GET /api/assets by @mattmillerai in #14014
• fix(assets): remove unused delete_content param from deleteAsset by @mattmillerai in #14241
• main: force cudnn.benchmark to false (CORE-284) by @rattus128 in #14390
• feat(assets): add job_ids filter to GET /api/assets by @mattmillerai in #13998
• feat(assets): include asset id in executed WebSocket message by @mattmillerai in #13862
• revert(assets): drop job_ids filter from GET /api/assets by @mattmillerai in #14408
• chore(openapi): sync shared API contract from cloud@e3c52ad by @comfy-pr-bot in #14406
• I don't think this actually works anymore. by @comfyanonymous in #14403
• ops: tolerate already force casted dynamic weight (CORE-288) by @rattus128 in #14410
• Improve context window resizing for SCAIL2 (CORE-286) by @drozbay in #14394
• Fix SCAIL-2 reference mask background convention by @kijai in #14415
• [Partner Nodes] fix(GPT Image): handle mismatched image sizes returned when size="auto" by @bigcat88 in #14414
• [Partner Nodes] fix(KlingTextToVideoNode): validation error for "kling-v2-master" model by @bigcat88 in #14418
• Make --enable-manager-legacy-ui imply --enable-manager by @Kosinkadink in #14421
• Don't crash when using flux kv cache with split batches. by @comfyanonymous in #14422
• Add Comfy-Usage-Source pass-through for API node requests by @robinjhuang in #14404
• [Partner Nodes] feat: enable Bria Replace Background node by @bigcat88 in #14397
• Fix potential dtype issue with ideogram 4. by @comfyanonymous in #14436
• add --high-ram option by @rattus128 in #14437
• [Partner Nodes] feat: add Runway Aleph2 node by @bigcat88 in #14306
• Use comfy kitchen apply rope in omnigen2 model. by @comfyanonymous in #14442
• feat(core-nodes): Add 10-bit video support by @bigcat88 in #14452
• Expose deploy_environment in /system_stats by @robinjhuang in #14402
• Fix nondeterministic video decode at unaligned widths (CORE-299) by @pollockjj in #14438
• [Partner Nodes] feat(Tripo3d): add new "Import 3D" node by @bigcat88 in #14466
• bump manager version to 4.2.2 by @ltdrdata in #14471
• This is already auto enabled by default. by @comfyanonymous in #14476
• Update embedded docs to v0.5.4 by @comfyui-wiki in #14478
• Comfy Aimdo 0.4.10 + Dynamic --reserve-vram + --vram-headroom by @rattus128 in #14480
• Update workflow templates to v0.10.0 by @comfyui-wiki in #14482
• [Partner Nodes] fix(api-nodes): always require "duration" to be specified for the SoniloTextToMusic node by @bigcat88 in #14484
• main: support fallback to aimdo 0.4.9 by @rattus128 in #14489
• Fix odd-height crash and edge bleed in unaligned-width image/video decode by @pollockjj in #14491
• chore: Update nodes titles by @alexisrolland in #14417

Full Changelog: v0.24.0...v0.25.0

What's Changed

• Some cast/dtype fixes for the birefnet and dino3 models. by @comfyanonymous in #14217
• Fix issue with triposplat preview and old offloading mode. by @comfyanonymous in #14218
• Do tripo dinov3 inference in fp32. by @comfyanonymous in #14221
• Fix uncaught OverflowError in Math Expression node for large int results by @vidigoat in #14214
• fix: Revert partner nodes category by @alexisrolland in #14229
• Radiance: support variant with nonzero txt_ids by @person4268 in #14206
• fix (MultiGPU): prevent freeze on manual abort when using MultiGPU CFG Split by @quasar-of-mikus in #14235
• PiD: Add SDXL and QwenImage by @kijai in #14240
• Remove old useless no comfy kitchen fallback. by @comfyanonymous in #14245
• comfy-aimdo: 0.4.8 by @rattus128 in #14244
• [Partner Nodes] feat: add Ideogram V4 node by @bigcat88 in #14261
• feat: Add Ideogram4 model support (CORE-208) by @kijai in #14259
• Mark DualModelGuider as experimental by @comfyanonymous in #14262
• Update workflow templates to v0.9.94 by @comfyui-wiki in #14263
• Memory usage factor for ideogram 4 on non dynamic vram. by @comfyanonymous in #14264

New Contributors

• @vidigoat made their first contribution in #14214
• @person4268 made their first contribution in #14206
• @quasar-of-mikus made their first contribution in #14235

Full Changelog: v0.23.0...v0.24.0

What's Changed

• feat: MediaPipe face detection (CORE-235) by @kijai in #14009
• Multi-threaded load of models from disk (big load time speedups & Offload to disk) (CORE-43,CORE-152,CORE-164,CORE-165,CORE-117) by @rattus128 in #13802
• Repo security stuff. by @comfyanonymous in #14019
• feat(openapi): add OAuth 2.1 + RFC 7591 DCR endpoints by @mattmillerai in #14026
• fix(openapi): correct POST /api/assets/import to importPublishedAssets by @mattmillerai in #14027
• openapi: remove cloud-only job_ids query param from GET /api/assets by @mattmillerai in #14016
• Update MediaPipe nodes to standardize with existing code base (CORE-242) by @alexisrolland in #14025
• chore: Dataset nodes clean-up (CORE-237) by @alexisrolland in #14002
• Update MoGe nodes display names, search aliases and descriptions by @alexisrolland in #14030
• fix(lanczos): correct dimension transposition for single-channel tensors (CORE-228) by @edoardocarmignani in #12679
• Fix reshaping lora application by @rattus128 in #14031
• [Partner Nodes] add OpenRouter LLM node by @bigcat88 in #14007
• Update workflow templates to v0.9.82 by @comfyui-wiki in #14034
• [Partner Nodes] add widget for auto upscaling for the ByteDance2Reference node by @bigcat88 in #14032
• Add backport release workflow. by @comfyanonymous in #14038
• Fix SIGPIPE false negative in Backport Release workflow validation by @Kosinkadink in #14041
• Add validation for source branch in backport workflow by @comfyanonymous in #14042
• Update backport workflow to use commit SHA input by @comfyanonymous in #14043
• fe: Consolidate warnings spam by @rattus128 in #13970
• Update Discord invite link in README.md by @comfyanonymous in #14045
• Adding in And, Or, and Not nodes. by @Pauan in #14004
• Move logic nodes into utils category by @alexisrolland in #14033
• [Partner Nodes] add new Rodin2.5 nodes by @bigcat88 in #14051
• openapi: align response declarations with implementation (5 endpoints) by @mattmillerai in #14058
• Delete the source branch after doing the backport. by @comfyanonymous in #14062
• openapi: rename 55 cloud-side operationIds to match runtime (PR A of 3) by @mattmillerai in #14060
• openapi: rename cloud-side response schemas to match runtime (PR D) by @mattmillerai in #14065
• openapi: add enum values + FeedbackRequest schema for cloud cutover (PR E) by @mattmillerai in #14070
• Remove dead code. by @comfyanonymous in #14072
• Bump comfyui-frontend-package to 1.44.19 by @comfy-pr-bot in #14074
• Fix --use-flash-attention ignored when xformers installed. by @comfyanonymous in #14083
• fix: correct description of where compiled FE files live by @robinjhuang in #14013
• openapi: fix invalid BillingStatus schema (object + enum hybrid) by @mattmillerai in #14071
• Aimdo 0.4.5 + Cold marking of pins by @rattus128 in #14088
• cache-ram: lower thresholds by @rattus128 in #14089
• Add colored logs by @Talmaj in #14036
• Save Image advanced node (CORE-32) by @comfyanonymous in #13850
• Update embedded docs to v0.5.1 by @comfyui-wiki in #14101
• Emit hash alongside asset_hash on all Asset responses by @mattmillerai in #13739
• Add new open-source model and built-in tool blueprints by @comfyui-wiki in #13980
• MultiGPU Work Units For Accelerated Sampling (CORE-184) by @Kosinkadink in #7063
• Remove some useless annotations imports. by @comfyanonymous in #14105
• multigpu: use unet_manual_cast for SelectModelDevice compute dtype by @Kosinkadink in #14108
• fix(LTX AV): pass correct timesteps to cross-attention AdaLN modulation by @izorinLightricks in #14097
• feat: Microsoft Lens support (CORE-248) by @kijai in #14077
• Lens: some cleanup by @kijai in #14112
• openapi: document QueueManageResponse body on POST /api/queue by @mattmillerai in #14117
• feat: Support NVIDIA PixelDiT and PiD (CORE-201) by @kijai in #14103
• Closer memory usage factors for PID by @comfyanonymous in #14123
• Add memory usage factor for lens model. by @comfyanonymous in #14124
• fix: Stop LTXVCropGuides leaving stray latent frames when guides share a start position (CORE-206) by @drozbay in #13882
• [Partner Nodes] feat: improve video references uploading for SeeDance 2 by @bigcat88 in #14098
• Update workflow templates to v0.9.85 by @comfyui-wiki in #14134
• [Partner Nodes] feat: add Krea2 Image nodes by @bigcat88 in #14130
• [V3] convert nodes_lt_upsampler nodes to V3 schema by @bigcat88 in #12423
• [Partner Nodes] feat: Beeble SwitchX nodes by @bigcat88 in #14137
• Support context window for PiD and fix lq_latent rounding by @kijai in #14136
• Disable sage attention in stable audio dit and VAE. by @comfyanonymous in #14148
• chore: Update nodes categories by @alexisrolland in #14145
• feat: add camera intrinsics fields to Load3DCamera info by @jtydhr88 in #14143
• Make Load3D model_file optional by adding "none" choice by @jtydhr88 in #13379
• [Partner Nodes] fix: add runtime check for SeeDance2 image inputs by @bigcat88 in #14152
• [Partner Nodes] chore: update the category of the Beeble nodes by @bigcat88 in #14156
• feat: add overwrite/increment to SaveImageTextDataSetToFolderNode by @zeerd in #13215
• Add unreviewed merge detector for SOC 2 compliance by @luke-mino-altherr in #14146
• Bump pyav package to fix some image loading issues. by @comfyanonymous in #14160
• refactor: drop rotation from Load3DCamera by @jtydhr88 in #14159
• Update and pin comfy-kitchen version to 0.2.9 by @comfyanonymous in #14161
• float: use CK stochastic rounding cuda kernel (CORE-200) by @rattus128 in #13971
• Update workflow templates to v0.9.91 by @comfyui-wiki in #14163
• feat: add model_info output to Load3D node by @jtydhr88 in #14144
• [Partner Nodes] fix: removed "beta" models versions from Grok nodes by @bigcat88 in #14170
• Fix background removal mask output shape by @kijai in #14171
• [Partner Nodes] feat: add new nodes for Tripo3D P1 model by @bigcat88 in #14155
• Remove useless code. by @comfyanonymous in #14178
• Speed up anima a bit on nvidia. by @comfyanonymous in #14181
• feat(io): add File3DPLY / File3DSPLAT / File3DSPZ / File3DKSPLAT types by @jtydhr88 in #14185
• Threaded Loader performance fixes / improvements (+ Aimdo 0.4.6) by @rattus128 in #14116
• Revert deprecation of non-dynamic smart memory (CORE-152 (revert)) by @rattus128 in #14183
• feat: add Preview3DAdvanced node by @jtydhr88 in #14175
• Update embedded docs to v0.5.2 by @comfyui-wiki in #14193
• Speed up ernie model by a bit on nvidia and use higher quality rope. by @comfyanonymous in #14192
• fix: enable XPU MultiGPU CFG Split on Windows (CFG >...

What's Changed

• Add test that each model has unique identifiers (CORE-134) by @Talmaj in #13654
• Make audio processing nodes handle None -inputs by @kijai in #13879
• Update workflow templates to v0.9.77 by @comfyui-wiki in #13895
• Add careers link to README and startup log by @robinjhuang in #13897
• Remove annoying message. by @comfyanonymous in #13899
• Include workflow_id in all execution WebSocket messages (CORE-198) by @christian-byrne in #13684
• Revert "Include workflow_id in all execution WebSocket messages (CORE-198)" by @comfyanonymous in #13901
• Create SECURITY.md. by @comfyanonymous in #13902
• Generalize frontend version warning to all comfy* requirements.txt entries by @christian-byrne in #13875
• Preserve noise_scale/shift across chained model_sampling patches by @kijai in #13892
• feat: Support MoGe (CORE-168) by @kijai in #13878
• Allow negative batch_index on ImageFromBatch and LatentFromBatch (CORE-195) by @drozbay in #13857
• Reduce LTX2.3 peak VRAM when guide_mask is in use (CORE-166) by @kijai in #13735
• Remove useless option and clarify what lowvram does. by @comfyanonymous in #13922
• feat: Add downscaled IC-LoRA support to LTXVAddGuide (CORE-102) by @drozbay in #13896
• Enhance README with application and cloud links by @comfyanonymous in #13936
• Log which quant ops are enabled/emulated. by @comfyanonymous in #13946
• chore: correct OOM memory summary format by @Apophis3158 in #13950
• Fix Qwen3.5 text generation with multi-image prompts by @kijai in #13943
• HiDream-O1: support area conditioning by @kijai in #13944
• [Partner Nodes] fix Opus 4.7 sending deprecated temperature parameter by @bigcat88 in #13955
• fix(FeatherMask): correct negative zero indexing for right/bottom feathering by @alvinttang in #12881
• implement/fix dynamic clip saving by @rattus128 in #13959
• [Partner Nodes] add ByteDance Seed LLM node by @bigcat88 in #13919
• Fix BiRefNet issue by @yousef-rafat in #13966
• feat: Add optional attention_mask input to LTXVAddGuide (CORE-220) by @drozbay in #13965
• Reduce min for Batch Image/Mask/Latent nodes from 2 to 1 by @Kosinkadink in #13721
• Initial work to make downscale_ratio_temporal work. by @comfyanonymous in #13972
• chore: Various QoL updates of nodes display names, descriptions and categories (CORE-190, CORE-191) by @alexisrolland in #13830
• Use temporal downscale to make empty audio latent nodes more reusable. by @comfyanonymous in #13975
• docs(openapi): deprecate /api/upload/mask in favor of /api/upload/image by @mattmillerai in #13968
• docs(openapi): remove top-level width/height fields on Asset schema by @mattmillerai in #13973
• Fix typo in ops.py by @NAM-MAN in #11925
• feat(openapi): add optional description field to workspace API key schemas by @mattmillerai in #13993
• docs(openapi): tighten workspace API key description field (BE-1004) by @mattmillerai in #13996
• Adding new StringFormat node by @Pauan in #13997
• Add spacial downscale ratio to empty ltxv latent. by @comfyanonymous in #13999
• fix: Hunyuan3D 2.1 batch size crashes in attention and forward pass by @Kivylius in #13699
• Support Stable Audio 3 model. by @comfyanonymous in #14010
• Update workflow templates to v0.9.79 by @comfyui-wiki in #14011

New Contributors

• @Apophis3158 made their first contribution in #13950
• @NAM-MAN made their first contribution in #11925
• @Pauan made their first contribution in #13997
• @Kivylius made their first contribution in #13699

Full Changelog: v0.21.1...v0.22.0

What's Changed

• [Partner Nodes] new Flux2ImageNode and GrokImageEditNodeV2 nodes by @bigcat88 in #13814
• [Partner Nodes] new ByteDanceSeedreamNodeV2 node with DynamicCombo and Autogrow by @bigcat88 in #13811
• [Partner Nodes] new OpenAI Image node with DynamicCombo and Autogrow by @bigcat88 in #13838
• model_patcher: Fix safetensors saving of fp8 (CORE-160) by @rattus128 in #13835
• Support anima TE lora kohya format. by @comfyanonymous in #13847
• feat: Support HiDream-O1-Image (CORE-187) by @kijai in #13817
• Fix dtype issue with hidream o1 by @comfyanonymous in #13849
• [Partner Nodes] fix Quiver nodes by @bigcat88 in #13851
• Mark deprecated cloud-runtime endpoints in spec by @millermedia in #13789
• Suppress false-positive Spectral lint on WebSocket endpoint by @millermedia in #13842
• Fix LTXV mid-video multi-frame guide alignment (CORE-129) by @drozbay in #13625
• Revert some breaking changes. by @comfyanonymous in #13861
• fix: create input directory if missing in LoadAudio define_schema by @angad777 in #13834
• Better Hidream O1 mem usage factor for non dynamic vram. by @comfyanonymous in #13864
• chore: update embedded docs to v0.5.0 by @comfyui-wiki in #13865
• Add Create Video to the essentials tab by @AustinMroz in #13863
• feat: Extend Save3D to save vertex colors and textures (CORE-189) by @kijai in #13824
• [Partner Nodes] add Claude LLM node by @bigcat88 in #13867
• chore: update workflow templates to v0.9.75 by @comfyui-wiki in #13877
• Fix VOID failing with RuntimeError: … by @Talmaj in #13873

New Contributors

• @angad777 made their first contribution in #13834

Full Changelog: v0.21.0...v0.21.1

What's Changed

• Read audio and video at the same time in video loader node. by @comfyanonymous in #13591
• fix: Handle un-inited meta-tensors in models (fixes a CPU TE crash) (CORE-67) by @rattus128 in #13578
• ci: dispatch tag push to Comfy Cloud by @millermedia in #13541
• Change the save 3d model node's filename prefix to 3d/ComfyUI (CORE-106) by @comfyui-wiki in #12826
• Fix SolidMask and MaskComposite device mismatch with --gpu-only (CORE-98) by @gschreiber in #13296
• Handle metadata rotation in pyav code. by @comfyanonymous in #13605
• comfy-aimdo v0.3.0 by @rattus128 in #13604
• Make pyav loading code handle tRNS PNG. by @comfyanonymous in #13607
• Use pyav to load images instead of pillow (CORE-95) by @comfyanonymous in #13594
• dynamicVRAM + --cache-ram 2 (CORE-117) by @rattus128 in #13603
• Reduce video tiny VAE peak VRAM and decode time (CORE-127) by @kijai in #13617
• Cogvideox by @Talmaj in #13402
• Add high quality preview support for Flux2 latents by @blepping in #13496
• Lower peak mem usage for 8 bit formats with pyav. by @comfyanonymous in #13626
• Fix SQLAlchemy version format in requirements.txt (CORE-105) by @Kosinkadink in #13547
• [Partner Nodes] allow custom resolutions for GPTImage2 node by @bigcat88 in #13631
• [Partner Nodes] ByteDance: virtual portrait library for regular images by @bigcat88 in #13638
• Load other jpeg formats without taking so much memory. by @comfyanonymous in #13642
• OneTrainer ERNIE LoRA support by @prookyon in #13640
• Update workflow templates to v0.9.65 by @comfyui-wiki in #13644
• Add alexisrolland and rattus128 as code owners by @Kosinkadink in #13648
• [Partner Nodes] always display the custom width and height in GPTImage2 by @bigcat88 in #13651
• Reformat models variable into multiline array (CORE-59) by @Talmaj in #13513
• chore(api-nodes): remove Moonvalley API nodes by @bigcat88 in #13659
• Update workflow templates to v0.9.66 by @comfyui-wiki in #13662
• chore(api-nodes): increase default timeout for partner API node tasks by @bigcat88 in #13663
• Remove IPEX (Intel Extension for Pytorch) support. by @simonlui in #13653
• SDPose: resize fix (CORE-146) by @kijai in #13656
• Adding kijai to codeowners by @alexisrolland in #13664
• List all the portable downloads in the README section. by @comfyanonymous in #13666
• Add script in AMD portable to launch with dynamic vram. by @comfyanonymous in #13667
• Implement block prefetch + Lora Async load + and adopt in LTX (Speedup!) (CORE-111) by @rattus128 in #13618
• Some cleanups to the load image node (CORE-145) by @comfyanonymous in #13677
• Update workflow templates to v0.9.68 by @comfyui-wiki in #13678
• [Partner Nodes] Topaz Astra 2 model by @bigcat88 in #13672
• feat: Gemma4 text generation support (CORE-30) by @kijai in #13376
• Update README (CORE-144) by @alexisrolland in #13679
• fix: Make ImageBlend and ImageCompositeMasked nodes handle images with different channel counts (CORE-103) by @alexisrolland in #13615
• Update ComfyUI screenshot in README (CORE-144) by @alexisrolland in #13683
• fix: gracefully handle port-in-use error on server startup by @luke-mino-altherr in #13001
• Batch broadcasting in JoinImageWithAlpha node (CORE-143) by @kijai in #13686
• Enable triton comfy kitchen via cli-arg by @silveroxides in #12730
• Refactor LoadImageMask to use LoadImage code (CORE-142) by @comfyanonymous in #13687
• PrimitiveInt: default control_after_generate widget to fixed (CORE-141) by @Kosinkadink in #13690
• fix: Make ColorTransfer node ref_image mandatory (CORE-140) by @alexisrolland in #13691
• Fix a1111 typo in extra_model_paths.yaml by @soof-golan in #2720
• fix: Proper memory estimation for frame interpolation (CORE-150) by @kijai in #13698
• prefetch: guard against no offload (CORE-149) by @rattus128 in #13703
• make control-net load order deterministic (CORE-139) by @rattus128 in #13701
• ops: handle multi-compute of the same weight (CORE-153) by @rattus128 in #13705
• chore: Update display names and categories (CORE-151) by @alexisrolland in #13693
• fix(spec): mark DeviceStats.index and NodeInfo.essentials_category as nullable by @millermedia in #13706
• spec: add workflow_id / workflow_version_id to PromptRequest with x-runtime tag by @millermedia in #13709
• feat: add generic --feature-flag CLI arg and --list-feature-flags registry by @Kosinkadink in #13685
• Add deploy environment header (Comfy-Env) to partner node API calls by @Kosinkadink in #13425
• Document core stable release frequency is now ~2 weeks. by @comfyanonymous in #13710
• feat: Auto-regressive video generation (CORE-25) by @Talmaj in #13082
• fix: SplitImageToTileList and ImageMergeTileList to use tile_height for vertical stride minimum by @alvinttang in #12882
• [Partner Nodes] add Luma UNI-1 model by @bigcat88 in #13614
• [Partner Nodes] add Gpt 5.5 and 5.5-pro LLM models by @bigcat88 in #13673
• Update workflow templates to v0.9.69 by @comfyui-wiki in #13714
• chore: Update display names and categories for text nodes (CORE-155) by @alexisrolland in #13712
• Fix Content-Disposition header missing attachment prefix by @themachinehf in #13093
• fix: make VAEDecodeAudio usable for LTX-2.x generated audio latents (CORE-157) by @drozbay in #13716
• Add Spectral lint CI gate for openapi.yaml by @millermedia in #13410
• fix(api-io): serialize MultiCombo multi_select as object config by @Kosinkadink in #13484
• feat: add cloud-specific fields to OSS openapi.yaml as nullable by @millermedia in #13623
• Add temporal_downscale_ratio to LatentFormat (CORE-147) by @drozbay in #13702
• feat: Context windows - add causal_window_fix to improve blending of context windows (CORE-100) by @drozbay in #13563
• First step of supporting save filenames without trailing _ by @comfyanonymous in #13722
• Update language options in nodes_ace.py by @iChristGit in #12578
• fix: return millisecond timestamps from get_file_info() by @luke-mino-altherr in #12996
• Void model - pass 1 & 2 (CORE-38) by @Talmaj in #13403
• Bump comfyui-frontend-package to 1.43.17 by @comfy-pr-bot in #13723
• Revert "Fix Content-Disposition header missing 'attachment;' prefix (#13093)" by @guill in #13733
• Throttle dynamic VRAM prepare logging by @kijai in #13704
• Add .comfy_environment to portable. by @comfyanonymous in #13746
• [Partner Nodes] add grok-imagine-image-quality model by @bigcat88 in #13725
• Update workflow templates to v0.9.72 by @comfyui-wiki in #13732
• [Partner Nodes] fix of price badge for Kling V3 model in the Motion Control by @bigcat88 in #13790
• [Partner Nodes] new NanoBanana2 node with DynamicCombo/Autogrow by @bigcat88 in https://githu...

What's Changed

• feat: SUPIR model support (CORE-17) by @kijai in #13250
• Some optimizations to make Ernie inference a bit faster. by @comfyanonymous in #13472
• fix: append directory type annotation to internal files endpoint (CORE-71) by @Abdulrehman-PIAIC80387 in #13305
• Add link to Intel portable to Readme. by @comfyanonymous in #13477
• fix: pin SQLAlchemy>=2.0 in requirements.txt (CORE-63) by @octo-patch in #13316
• Refactor io to IO in nodes_ace.py by @comfyanonymous in #13485
• Bump comfyui-frontend-package to 1.42.12 by @comfy-pr-bot in #13489
• Make the ltx audio vae more native (CORE-76) by @comfyanonymous in #13486
• [Partner Nodes] add optional auto-downscaling of videos for ByteDance 2 nodes by @bigcat88 in #13465
• Support standalone LTXV audio VAEs by @kijai in #13499
• [Partner Nodes] added 4K resolution for Veo models; added Veo 3 Lite model by @bigcat88 in #13330
• Bump comfyui-frontend-package to 1.42.14 by @comfy-pr-bot in #13493
• Add gpt-image-2 as version option by @AustinMroz in #13501
• Allow logging in comfy app files. by @comfyanonymous in #13505
• Update workflow templates to v0.9.59 by @comfyui-wiki in #13507
• fix(veo): reject 4K resolution for veo-3.0 models by @millermedia in #13504
• feat: RIFE and FILM frame interpolation model support (CORE-29) by @kijai in #13258
• fix: use Parameter assignment for Stable_Zero123 cc_projection weights by @octo-patch in #13518
• Derive InterruptProcessingException from BaseException by @blepping in #13523
• bump manager version to 4.2.1 by @ltdrdata in #13516
• ModelPatcherDynamic: force cast stray weights on comfy layers by @rattus128 in #13487
• Update logging level for invalid version format by @comfyanonymous in #13526
• [Partner Nodes] add SD2 real human support by @bigcat88 in #13509
• feat: SAM (segment anything) 3.1 support (CORE-34) by @kijai in #13408
• [Partner Nodes] GPTImage: fix price badges, add new resolutions by @bigcat88 in #13519
• Update workflow templates to v0.9.61 by @comfyui-wiki in #13533
• Update embedded docs to v0.4.4 by @comfyui-wiki in #13535
• [Partner Nodes] add 4K resolution to Kling nodes by @bigcat88 in #13536
• Fix LTXV Reference Audio node by @kijai in #13531
• comfy-aimdo 0.2.14: Hotfix async allocator estimations by @rattus128 in #13534
• Disable sageattention for SAM3 by @kijai in #13529
• execution: Add anti-cycle validation by @rattus128 in #13169
• Update workflow templates to v0.9.62 by @comfyui-wiki in #13539
• Print more tensor values in the preview any node. by @comfyanonymous in #13544
• fix: use textureSize instead of u_resolution for texel size in blur/sharpen shaders by @jtydhr88 in #13347
• range type by @jtydhr88 in #13322
• Add OpenAPI 3.1 specification for ComfyUI API by @millermedia in #13397
• chore(api-nodes): add upcoming-deprecation notice to Sora nodes by @bigcat88 in #13549
• Properly load higher bit depth videos. by @comfyanonymous in #13542
• Bump comfyui-frontend-package to 1.42.15 by @comfy-pr-bot in #13556
• Support loading the alpha channel of videos. by @comfyanonymous in #13564
• Add new ComfyUI blueprints and fix subgraph naming by @comfyui-wiki in #13371
• Make EmptySD3LatentImage node use intermediate dtype. by @comfyanonymous in #13577
• [Partner Nodes] HappyHorse model by @bigcat88 in #13582
• Update workflow templates to v0.9.63 by @comfyui-wiki in #13586

New Contributors

• @Abdulrehman-PIAIC80387 made their first contribution in #13305
• @octo-patch made their first contribution in #13316

Full Changelog: v0.19.3...v0.20.1**Full Changelog**: v0.20.0...v0.20.1