Commit Graph

  • 7d512fa9c3 feat(isolation): core infrastructure and pyisolate integration John Pollock 2026-04-07 05:58:03 -0500
  • 2beca418ad
    Fix noise/latent tensor mismatch when latent is nested but noise is not Dustin 2026-04-07 06:07:26 -0400
  • 05eaceafa1 Cleanup, video fixes kijai 2026-04-07 12:37:29 +0300
  • ee600a3cce
    Merge branch 'master' into pysssss/angle-glsl pythongosssss 2026-04-07 10:16:31 +0100
  • b615af1c65
    Add support for small flux.2 decoder (#13314) huemin 2026-04-07 00:44:18 -0700
  • 40862c0776
    Support Ace Step 1.5 XL model. (#13317) comfyanonymous 2026-04-07 00:13:47 -0700
  • e3bf317153 Support Ace Step 1.5 XL model. comfyanonymous 2026-04-07 03:06:53 -0400
  • d656ca334a fix: pin SQLAlchemy>=2.0 in requirements.txt (fixes #13036) Octopus 2026-04-07 13:32:52 +0800
  • 50076f3439
    format blueprint (#13315) Terry Jia 2026-04-06 23:33:55 -0400
  • ea71c97e3d
    Merge branch 'master' into format-blueprint guill 2026-04-06 20:30:06 -0700
  • 3abd8c64fa format blueprint Terry Jia 2026-04-06 22:34:00 -0400
  • a0a33ca618 Add support for small flux.2 decoder huemin-art 2026-04-07 01:46:45 +0000
  • 61c2387436
    Ace step empty latent nodes follow intermediate dtype. (#13313) comfyanonymous 2026-04-06 18:12:16 -0700
  • 0fecf6f60c
    Ace step empty latent nodes follow intermediate dtype. comfyanonymous 2026-04-06 18:04:54 -0700
  • 93e8635110 parity with reference implementation kijai 2026-04-07 01:15:04 +0300
  • 6e831a1ab3 feat: add proxyWidgetSelector to subgraph blueprints proxy-widget-selector Terry Jia 2026-04-03 18:11:32 -0400
  • 7083484a48
    image histogram node (#13153) Terry Jia 2026-04-06 17:54:02 -0400
  • 3a061f4bbf LTX2 context windows - Cleanup: Simplify IndexListContextHandler standard execute path ozbayb 2026-04-06 15:13:46 -0600
  • 1cd2730b25
    Merge branch 'master' into deepme987/auto-register-node-replacements-json Deep Mehta 2026-04-06 13:13:42 -0700
  • ffec484ca6
    Merge branch 'master' into image-histogram guill 2026-04-06 12:56:13 -0700
  • ad89c87197
    Merge 19fd53dbb3 into 4b1444fc7a thetacola 2026-04-06 18:53:48 +0000
  • 874690c01c LTX2 context windows - Refactor guide logic from context_windows into LTXAV model hooks ozbayb 2026-04-06 11:44:14 -0600
  • a384f87a8d
    Merge 4202093225 into 4b1444fc7a boriselec 2026-04-06 17:41:39 +0000
  • 350237618d LTX2 context windows - Cleanup: Remove model specific code from BaseModel. Older LTXV model's guides + context_windows will need to be re-implemented but outside the scope of LTX2 changes ozbayb 2026-04-06 09:11:35 -0600
  • 3660533f83 LTX2 context windows - Cleanup: latent_start value is required for context windows with guides ozbayb 2026-04-06 08:48:51 -0600
  • 71712472f5 LTX2 context windows - Ensure that inplace latent images are retained properly with the retain index list ozbayb 2026-04-05 14:32:06 -0600
  • 9566c18ced LTX2 context windows - Fix crash when a window doesn't have a guide index ozbayb 2026-04-05 14:30:54 -0600
  • ef61ddfaed Fix freenoise application for LTXAV context windows, fix audio mapping to context windows ozbayb 2026-03-27 11:24:40 -0600
  • 115dbb69d1 LTX2 context windows part 3 - Generalize guide splitting to windows ozbayb 2026-03-27 11:23:43 -0600
  • 941d50e777 LTX2 context windows part 2b - Calculate guide parameters in model code, refactor ozbayb 2026-03-24 11:43:42 -0600
  • 56de390c25 LTX2 context windows part 2 - Guide aware processing ozbayb 2026-03-23 13:50:28 -0600
  • 5bfe660b7c Test implementation for LTX2 context windows ozbayb 2026-03-22 07:46:45 -0600
  • 498fc2e39a
    Merge branch 'master' into master yt-koike 2026-04-06 22:44:32 +0900
  • d38e5b847b
    Update app/model_manager.py yt-koike 2026-04-06 22:44:03 +0900
  • ffd4a002ab
    Update app/model_manager.py yt-koike 2026-04-06 22:43:20 +0900
  • 447b15b9a0
    Update app/model_manager.py yt-koike 2026-04-06 22:42:58 +0900
  • 595ae36e91
    Merge 14700174a5 into 4b1444fc7a newsletternewsletter 2026-04-06 12:17:41 +0000
  • 14e7ca7dc8
    fix: skip NaN keypoints in SDPoseDrawKeypoints to prevent ValueError (#13149) AbdulRehman 2026-04-06 14:51:29 +0500
  • 5763829b55
    Merge 3f56e64959 into 4b1444fc7a salt 2026-04-06 09:31:07 +0000
  • 4cdc631b62
    Merge f499b568b0 into 4b1444fc7a Lex Darlog 2026-04-06 09:24:26 +0000
  • 35def5cada
    fix: append directory type annotation to internal files endpoint response (#13078) AbdulRehman 2026-04-06 13:42:05 +0500
  • b8bb5427f9 fix: make attention_mask optional in LTXBaseModel.forward (fixes #13299) Octopus 2026-04-06 13:17:51 +0800
  • 3dc870f510
    Merge 6a2ceb470b into 4b1444fc7a Mark Ledford 2026-04-06 02:10:24 +0000
  • cf43a3a63e Use commonpath for path traversal check jakelodwick 2026-04-05 19:16:52 -0600
  • 8088b347d0 Fix torch.load missing weights_only in LoadTrainingDataset jakelodwick 2026-04-05 19:13:08 -0600
  • 4b1444fc7a
    Update README.md with new frontend release cycle. (#13301) comfyanonymous 2026-04-05 16:37:27 -0700
  • 94331804f2
    Update README.md with new frontend release cycle. comfyanonymous 2026-04-05 16:36:58 -0700
  • 7f71575235 qwen35: support flash linear attention. omarom 2026-04-05 14:15:44 +0000
  • 2729bd0dc6 Fix SolidMask and MaskComposite device mismatch with --gpu-only gschreiber 2026-04-05 12:46:56 +0000
  • 209d00c3b2 fix: MPS synchronization and --lowvram/--novram support spn 2026-04-05 12:37:14 +0200
  • 987dce1db9
    Merge branch 'Comfy-Org:master' into fix/qwen35-partial-load Silver 2026-04-05 12:27:19 +0200
  • 9e8e8e4a96 Fix device consistency for manual parameters and rotary embeddings in Qwen3.5 silveroxides 2026-04-05 12:21:31 +0200
  • d6756e5c97 Fix device mismatch. 1. In apply_rope, move the RoPE frequency tensors (cos, sin, nsin) to the device of the input tensor xq. 2. In Qwen35VisionModel.forward, move position_embeddings to x.device. silveroxides 2026-04-05 11:57:21 +0200
  • d2c287a368
    Merge branch 'Comfy-Org:master' into master austin1997 2026-04-05 17:27:07 +0800
  • 7e1e4a182d
    Merge pull request #1 from austin1997/add-grok-api-key-support-9519594855872412057 austin1997 2026-04-05 17:26:37 +0800
  • d546002990 feat(grok): sanitize user-provided xAI API key google-labs-jules[bot] 2026-04-05 09:15:48 +0000
  • f14b55f0d7 fix(grok): sanitize user-provided xAI API key google-labs-jules[bot] 2026-04-05 09:01:28 +0000
  • 679e3bbfc8
    Merge branch 'master' into nilay blackLion 2026-04-05 11:06:28 +0530
  • 40f8974443 initial setup nygupta 2026-04-05 11:03:13 +0530
  • 8cbbea8f6a
    chore: update workflow templates to v0.9.44 (#13290) Daxiong (Lin) 2026-04-05 13:31:11 +0800
  • 76fe91c906 fix: raise clear error when non-LLM model is used with TextGenerate node (fixes #13286) Octopus 2026-04-05 13:21:53 +0800
  • 6648f31c55 fix: raise clear error when non-LLM model is used with TextGenerate node (fixes #13286) Octopus 2026-04-05 13:12:39 +0800
  • 3aed5ae54c chore: update workflow templates to v0.9.44 ComfyUI Wiki 2026-04-05 13:01:31 +0800
  • cc96b933fc feat(grok): add support for official xAI API keys in Grok nodes google-labs-jules[bot] 2026-04-05 02:52:16 +0000
  • 13917b3880
    Nightly Nvidia pytorch is now cu132 (#13288) comfyanonymous 2026-04-04 16:02:47 -0700
  • 59d6601d32
    Nightly Nvidia pytorch is now cu132 comfyanonymous 2026-04-04 16:00:24 -0700
  • 6447250bd6 Add HY-OmniWeave support for HunyuanVideo 1.5 Codex 2026-04-04 22:03:24 +0000
  • 33b967d30a fix: MPS memory reporting and cache synchronization spn 2026-04-04 21:14:58 +0200
  • 49fef1697c
    Replace empty tensor initialization with zeros azazeal04 2026-04-04 20:21:27 +0200
  • 5d119f0532
    Implement allow_fp16 configuration in model detection azazeal04 2026-04-04 20:09:10 +0200
  • b2e08bb963
    Avoid overwriting existing keys in state_dict azazeal04 2026-04-04 20:05:17 +0200
  • 6965b83a2a
    Fix condition and update transformer_options handling azazeal04 2026-04-04 19:57:37 +0200
  • 479536e34d
    Prevent overwriting existing keys in state_dict azazeal04 2026-04-04 19:45:05 +0200
  • a73ac772b5
    Fix key mapping for t_embedder compatibility azazeal04 2026-04-04 19:42:55 +0200
  • 5f041f4e5e
    Refactor time embedding and projection layers azazeal04 2026-04-04 19:38:55 +0200
  • de2ff57f3c
    Implement _compute_twinflow_adaln method azazeal04 2026-04-04 18:49:19 +0200
  • 4fe75dcce8
    Add configuration for TwinFlow-Z-Image model azazeal04 2026-04-04 17:23:46 +0200
  • 3cbd1d5f71 Better RAM usage, reduce FILM VRAM peak kijai 2026-04-04 17:28:37 +0300
  • fec629ba3d
    Implement twinflow_z_image function azazeal04 2026-04-04 16:09:22 +0200
  • e978c61a79
    Implement twinflow_z_image_key_mapping function azazeal04 2026-04-04 16:06:41 +0200
  • b76510b549
    Add TwinFlow_Z_Image model to supported models azazeal04 2026-04-04 16:03:52 +0200
  • dc413071ff
    Add functions to check and extract TwinFlow-Z-Image config azazeal04 2026-04-04 15:59:50 +0200
  • 84c983a779
    Add detection for TwinFlow-Z-Image checkpoints azazeal04 2026-04-04 15:57:19 +0200
  • 21ed4a2242
    Add TwinFlow_Z_Image class to model_base.py azazeal04 2026-04-04 15:55:09 +0200
  • 02b5222dcc
    Add twinflow_z_image_lora_to_diffusers function azazeal04 2026-04-04 15:51:35 +0200
  • 00c4e72a97
    Implement TwinFlow-Z-Image LoRA compatibility azazeal04 2026-04-04 15:49:32 +0200
  • 257c5312d9 Also support FILM kijai 2026-04-04 16:09:34 +0300
  • c254fa10d3 Add inference scaling nodes with quality monitoring during generation kasailab 2025-12-03 11:26:04 +0900
  • 3d51d63490 fix: suppress false-positive warnings when loading whisper audio encoder Octopus 2026-04-04 13:45:54 +0800
  • f21f6b2212
    Add portable release for intel XPU. (#13272) comfyanonymous 2026-04-03 12:29:06 -0700
  • f34b175da9
    Add portable release for intel XPU. comfyanonymous 2026-04-03 15:23:30 -0400
  • 8705579536 fix: make comfy_aimdo import optional for AMD/ROCm GPU support Octopus 2026-04-03 16:00:12 +0800
  • 7782171a27 ComfyUi v0.18.5 v0.18.5 release/v0.18.5 comfyanonymous 2026-04-03 03:12:59 -0400
  • c6e15cb366 Revert "Fix some fp8 scaled checkpoints no longer working. (#13239)" v0.18.4 release/v0.18.4 comfyanonymous 2026-04-03 03:09:58 -0400
  • 6aba6bd435 ComfyUI version 0.18.4 comfyanonymous 2026-04-03 03:06:35 -0400
  • dedec11116 Update template to 0.9.43 (#13265) Daxiong (Lin) 2026-04-03 14:52:10 +0800
  • b818eacb0b feat(api-nodes): new Partner nodes for Wan2.7 (#13264) Alexander Piskun 2026-04-03 09:51:47 +0300
  • 4504ee718a Fix some fp8 scaled checkpoints no longer working. (#13239) comfyanonymous 2026-03-31 14:27:17 -0700
  • 94ab69f609 Fix/tweak pinned memory accounting (#13221) rattus 2026-03-29 16:43:24 -0700
  • eb0686bbb6
    Update template to 0.9.43 (#13265) Daxiong (Lin) 2026-04-03 14:52:10 +0800