{"status":"success","data":{"lastUpdated":"2026-09-14","optionalModelPacks":{"pixal3d":{"displayName":"Pixal3D image to 3D","description":"Adds prompt-free 3D reconstruction from a single image or from a front view plus up to three orbit views, with BiRefNet subject isolation, high-detail geometry, PBR textures, and GLB export.","workflowIds":["pixal3d_int8_i23d","pixal3d_multiview_int8_i23d"],"downloadGroupIds":["pixal3d-core","pixal3d-multiview","pixal3d-birefnet"],"downloadBytes":15534998945,"minVramGB":30,"imageTracks":["cu13"],"localComfyOnly":true},"birefnet":{"displayName":"BiRefNet background removal","description":"Adds prompt-free subject cutout and mask generation at the source image's own dimensions.","workflowIds":["birefnet_image_background_removal_fp16"],"downloadGroupIds":["birefnet-background-removal"],"downloadBytes":444473596,"minVramGB":8,"imageTracks":["cu13"],"localComfyOnly":true},"sam3":{"displayName":"SAM 3 interactive segmentation","description":"Adds interactive object selection for masks and choose-your-own-adventure scene branching.","workflowIds":["sam3_image_segment_bf16"],"downloadGroupIds":["sam3-image-segment"],"downloadBytes":3450069594,"minVramGB":8,"imageTracks":["cu13"],"localComfyOnly":true}},"downloads":{"wan_v2.2-14b-fp8_shared":{"description":"Models shared by all Wan 2.2 14B fp8 workflows","files":[{"tensorFile":"wan_2.1_vae.safetensors","destinationFolder":"models/vae","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/vae/wan_2.1_vae.safetensors","bytes":253815318,"sha256":"2fc39d31359a4b0a64f55876d8ff7fa8d780956ae2cb13463b0223e15148976b"},{"tensorFile":"umt5_xxl_fp8_e4m3fn_scaled.safetensors","destinationFolder":"models/text_encoders","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/text_encoders/umt5_xxl_fp8_e4m3fn_scaled.safetensors","bytes":6735906897,"bytesExact":true,"sha256":"c3355d30191f1f066b26d93fba017ae9809dce6c627dda5f6a66eaa651204f68"}]},"wan_v2.2-14b-fp8_t2v":{"description":"Text-to-Video base workflow models","files":[{"tensorFile":"wan2.2_t2v_high_noise_14B_fp8_scaled.safetensors","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/diffusion_models/wan2.2_t2v_high_noise_14B_fp8_scaled.safetensors","bytes":14293923632,"sha256":"cad711ae211c8b23455ec68cd6a190a33a3d874234a77eb57266d73f8f0e6c9f"},{"tensorFile":"wan2.2_t2v_low_noise_14B_fp8_scaled.safetensors","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/diffusion_models/wan2.2_t2v_low_noise_14B_fp8_scaled.safetensors","bytes":14293923632,"sha256":"e71b96d7c82e638694c5e7fb98fac4bfb0e4ddc5fbbb4b1df40da8f0f1278a97"}]},"wan_v2.2-14b-fp8_t2v_lightx2v":{"description":"Lightx2v speed LoRAs for Text-to-Video workflow","files":[{"tensorFile":"wan2.2_t2v_lightx2v_4steps_lora_v1.1_high_noise.safetensors","description":"Lightx2v Speed LoRA for T2V high noise model","destinationFolder":"models/loras","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/loras/wan2.2_t2v_lightx2v_4steps_lora_v1.1_high_noise.safetensors","bytes":1525000000,"sha256":"698321cb86bd30c4af06c9b84e656a1048c8cb54e06d50694536fb5de37fde41"},{"tensorFile":"wan2.2_t2v_lightx2v_4steps_lora_v1.1_low_noise.safetensors","description":"Lightx2v Speed LoRA for T2V low noise model","destinationFolder":"models/loras","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/loras/wan2.2_t2v_lightx2v_4steps_lora_v1.1_low_noise.safetensors","bytes":1525000000,"sha256":"ec95216e614b3c132c11bfb387b11feedf62163150ccc9068bca8a189771e75a"}]},"wan_v2.2-14b-fp8_i2v":{"description":"Image-to-Video base workflow models","files":[{"tensorFile":"wan2.2_i2v_high_noise_14B_fp8_scaled.safetensors","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/diffusion_models/wan2.2_i2v_high_noise_14B_fp8_scaled.safetensors","bytes":14294742832,"sha256":"6122e79d55e0f235698d11d657f3b196c5273c830da00b2b013c5a048d5e6a42"},{"tensorFile":"wan2.2_i2v_low_noise_14B_fp8_scaled.safetensors","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/diffusion_models/wan2.2_i2v_low_noise_14B_fp8_scaled.safetensors","bytes":14294742832,"sha256":"5471a457b6ac404202a5fbe6c11595a3d5641fc766b00f38763f72303fffc21e"},{"tensorFile":"clip_vision_h.safetensors","destinationFolder":"models/clip_vision","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/clip_vision/clip_vision_h.safetensors","bytes":1264219396,"bytesExact":true,"sha256":"64a7ef761bfccbadbaa3da77366aac4185a6c58fa5de5f589b42a65bcc21f161"}]},"wan_v2.2-14b-fp8_i2v_lightx2v":{"description":"Lightx2v speed LoRAs for Image-to-Video workflow","files":[{"tensorFile":"wan2.2_i2v_lightx2v_4steps_lora_v1_high_noise.safetensors","description":"Lightx2v Speed LoRA for I2V high noise model","destinationFolder":"models/loras","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/loras/wan2.2_i2v_lightx2v_4steps_lora_v1_high_noise.safetensors","bytes":1525000000,"sha256":"d176c808d6fc461999b68e321efcb7501b20b8c3797523ed0df14f7d1deff11e"},{"tensorFile":"wan2.2_i2v_lightx2v_4steps_lora_v1_low_noise.safetensors","description":"Lightx2v Speed LoRA for I2V low noise model","destinationFolder":"models/loras","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/loras/wan2.2_i2v_lightx2v_4steps_lora_v1_low_noise.safetensors","bytes":1525000000,"sha256":"024f21de095bc8fad9809ded3e9e49a2e170dcf27075da8145ba7d60d8aab7f9"}]},"wan_v2.2-14b-fp8_s2v":{"description":"Sound-to-Video base workflow models","files":[{"tensorFile":"wan2.2_s2v_14B_fp8_scaled.safetensors","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/diffusion_models/wan2.2_s2v_14B_fp8_scaled.safetensors","bytes":16394832474,"sha256":"140e75af5534ac3d91e710d9df756f7032addd64b341ba2c1c70e3e6da9aa216"},{"tensorFile":"wav2vec2_large_english_fp16.safetensors","description":"Audio encoder for Sound-to-Video workflow","destinationFolder":"models/audio_encoders","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/audio_encoders/wav2vec2_large_english_fp16.safetensors","bytes":661000000,"sha256":"f0017a43ea57ef6b3d4866be607844bbd8cada6d30966f7d70044ed0d63d3f9e"}]},"wan_v2.2-14b-fp8_animate":{"description":"Animation base workflow models (includes DWPose for pose detection)","files":[{"tensorFile":"Wan2_2-Animate-14B_fp8_e4m3fn_scaled_KJ.safetensors","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/diffusion_models/Wan2_2-Animate-14B_fp8_e4m3fn_scaled_KJ.safetensors","bytes":18401760586,"sha256":"2936b31473a967e7a429a6646bba60e7862d0938e178b58b2a140f391dd5b8e6"},{"tensorFile":"yolox_l.onnx","description":"DWPose YOLOX model for pose detection","destinationFolder":"custom_nodes/comfyui_controlnet_aux/ckpts/yzd-v/DWPose","tensorDownload":"https://cdn.sogni.ai/ComfyUI/custom_nodes/comfyui_controlnet_aux/ckpts/yzd-v/DWPose/yolox_l.onnx","bytes":216746733,"sha256":"7860ae79de6c89a3c1eb72ae9a2756c0ccfbe04b7791bb5880afabd97855a411"},{"tensorFile":"WanAnimate_relight_lora_fp16.safetensors","description":"Relight LoRA for Animate workflow","destinationFolder":"models/loras","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/loras/WanAnimate_relight_lora_fp16.safetensors","bytes":630000000,"sha256":"fc646c74c73f4b251f5fd9bc440ef21b03b27305f499966c68b2b3aa31498561"}]},"controlnet-aux-depth-anything-v2-small":{"description":"Pinned offline Depth Anything V2 Small auxiliary for LTX depth control","files":[{"tensorFile":"depth_anything_v2_vits.pth","description":"Depth Anything V2 Small (ViT-S) depth estimator","destinationFolder":"custom_nodes/comfyui_controlnet_aux/ckpts/depth-anything/Depth-Anything-V2-Small","tensorDownload":"https://cdn.sogni.ai/ComfyUI/custom_nodes/comfyui_controlnet_aux/ckpts/depth-anything/Depth-Anything-V2-Small/depth_anything_v2_vits.pth","bytes":99218434,"bytesExact":true,"sha256":"715fade13be8f229f8a70cc02066f656f2423a59effd0579197bbf57860e1378"}]},"controlnet-aux-yolox-torchscript":{"description":"Pinned offline YOLOX TorchScript detector for LTX pose control","files":[{"tensorFile":"yolox_l.torchscript.pt","description":"YOLOX Large TorchScript person detector","destinationFolder":"custom_nodes/comfyui_controlnet_aux/ckpts/hr16/yolox-onnx","tensorDownload":"https://cdn.sogni.ai/ComfyUI/custom_nodes/comfyui_controlnet_aux/ckpts/hr16/yolox-onnx/yolox_l.torchscript.pt","bytes":217697649,"bytesExact":true,"sha256":"80bc14b13c260c24b3014cd42c02994bf52296ab8fa2d80a60b6afe08c93ef42"}]},"controlnet-aux-dwpose-torchscript":{"description":"Pinned offline DWPose TorchScript estimator shared by WAN Animate and LTX pose control","files":[{"tensorFile":"dw-ll_ucoco_384_bs5.torchscript.pt","description":"DWPose TorchScript BatchSize5 pose estimator","destinationFolder":"custom_nodes/comfyui_controlnet_aux/ckpts/hr16/DWPose-TorchScript-BatchSize5","tensorDownload":"https://cdn.sogni.ai/ComfyUI/custom_nodes/comfyui_controlnet_aux/ckpts/hr16/DWPose-TorchScript-BatchSize5/dw-ll_ucoco_384_bs5.torchscript.pt","bytes":135059124,"bytesExact":true,"sha256":"d86a0b2b59fddc0901a7076e9f59c9f8602602133ed72511c693fd11eea23d91"}]},"wan_v2.2-14b-fp8_animate_lightx2v":{"description":"Lightx2v speed LoRA for Animation workflow","files":[{"tensorFile":"lightx2v_I2V_14B_480p_cfg_step_distill_rank64_bf16.safetensors","description":"Lightx2v Speed LoRA for Animate workflow","destinationFolder":"models/loras","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/loras/lightx2v_I2V_14B_480p_cfg_step_distill_rank64_bf16.safetensors","bytes":738005744,"bytesExact":true,"sha256":"85c4a61c30e0497aa44b91d93a893b624708461a56fe5485183b28fa07e2dfb3"}]},"wan_v2.2-14b-fp8_animate-replace_lightx2v":{"description":"SAM2 model required only for animate-replace workflow","files":[{"tensorFile":"sam2_hiera_base_plus.safetensors","destinationFolder":"models/sam2","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/sam2/sam2_hiera_base_plus.safetensors","bytes":323407992,"sha256":"fa02d9028dcc4859c191f1d3f1ca1f7eefdb85f3b5e746c9ad738f322f3e89e2"}]},"z_image_turbo_bf16":{"description":"Z-Image Turbo text-to-image model","files":[{"tensorFile":"z_image_turbo_bf16.safetensors","description":"Z-Image Turbo Diffusion Model","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/diffusion_models/z_image_turbo_bf16.safetensors","bytes":12309866400,"sha256":"2407613050b809ffdff18a4ac99af83ea6b95443ecebdf80e064a79c825574a6"},{"tensorFile":"qwen_3_4b.safetensors","description":"Z-Image Turbo Text Encoder (Qwen 3 4B)","destinationFolder":"models/text_encoders","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/text_encoders/qwen_3_4b.safetensors","bytes":8044982048,"sha256":"6c671498573ac2f7a5501502ccce8d2b08ea6ca2f661c458e708f36b36edfc5a"},{"tensorFile":"ae.safetensors","description":"Z-Image Turbo VAE","destinationFolder":"models/vae","tensorDownload":"https://cdn.sogni.ai/vae/ae.safetensors","bytes":335304388,"sha256":"afc8e28272cd15db3919bacdb6918ce9c1ed22e96cb12c4d5ed0fba823529e38"}]},"krea2_turbo_fp8_scaled":{"description":"Krea 2 Turbo text-to-image model (distilled, 16GB+ VRAM)","minVramGB":16,"files":[{"tensorFile":"krea2_turbo_fp8_scaled.safetensors","description":"Krea 2 Turbo Diffusion Model (FP8 scaled)","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/diffusion_models/krea2_turbo_fp8_scaled.safetensors","bytes":13141730784,"sha256":"eb4dd8c612cfd10f64f25b057e6e6bbcb5737c94a7372177e456dbf7579502f1"},{"tensorFile":"qwen3vl_4b_fp8_scaled.safetensors","description":"Krea 2 Text Encoder (Qwen3-VL-4B FP8 scaled)","destinationFolder":"models/text_encoders","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/text_encoders/qwen3vl_4b_fp8_scaled.safetensors","bytes":5242467968,"sha256":"54bd5144df0bbc25dd6ccadfcb826b521445a1b06ae5a42570bdd2974ca87094"},{"tensorFile":"qwen_image_vae.safetensors","description":"Qwen Image VAE (shared with Qwen Image workflows)","destinationFolder":"models/vae","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/vae/qwen_image_vae.safetensors","bytes":253806246,"sha256":"a70580f0213e67967ee9c95f05bb400e8fb08307e017a924bf3441223e023d1f"},{"tensorFile":"krea2RealVae_v10.safetensors","description":"Krea 2 Real VAE v1.0","destinationFolder":"models/vae","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/vae/krea2RealVae_v10.safetensors","bytes":507591212,"sha256":"0dbbe0baeca04c2b98d2f3809c6f595608939809c88b695ba971368f17c874b8"},{"tensorFile":"wan_2.1_vae.safetensors","description":"WAN 2.1 VAE (popular alternate Krea 2 VAE)","destinationFolder":"models/vae","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/vae/wan_2.1_vae.safetensors","bytes":253815318,"sha256":"2fc39d31359a4b0a64f55876d8ff7fa8d780956ae2cb13463b0223e15148976b"}]},"krea2_identity_edit_v1_2":{"description":"Instruction-based, identity-preserving image editing for Krea 2. Give it an image and a plain-language instruction; it edits while preserving what you didn't ask to change, including the person. Source: Hugging Face @ conradlocke, https://huggingface.co/conradlocke/krea2-identity-edit","minVramGB":16,"files":[{"tensorFile":"krea2_identity_edit_v1_2.safetensors","description":"Krea 2 Identity Edit v1.2 LoRA from conradlocke/krea2-identity-edit","destinationFolder":"models/loras","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/loras/krea2_identity_edit_v1_2.safetensors","bytes":1828256432,"sha256":"6adf9a69cc9502d286db7b69964d37da7e9cfe4b05b4d004bc275f087d3fd3cf"}]},"qwen_image_edit_2511_fp8":{"description":"Qwen Image Edit 2511 image editing model","files":[{"tensorFile":"qwen_image_edit_2511_fp8mixed.safetensors","description":"Qwen Image Edit 2511 Diffusion Model (FP8)","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/diffusion_models/qwen_image_edit_2511_fp8mixed.safetensors","bytes":20533762817,"bytesExact":true,"sha256":"c9fdc158e46d3b61ef75f21ae866ca2fe808bf4a53643120d1c1e87c19280a4e"},{"tensorFile":"qwen_2.5_vl_7b_fp8_scaled.safetensors","description":"Qwen Image Edit Text Encoder (Qwen 2.5 VL 7B FP8)","destinationFolder":"models/text_encoders","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/text_encoders/qwen_2.5_vl_7b_fp8_scaled.safetensors","bytes":9384670680,"sha256":"cb5636d852a0ea6a9075ab1bef496c0db7aef13c02350571e388aea959c5c0b4"},{"tensorFile":"qwen_image_vae.safetensors","description":"Qwen Image VAE","destinationFolder":"models/vae","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/vae/qwen_image_vae.safetensors","bytes":253806246,"sha256":"a70580f0213e67967ee9c95f05bb400e8fb08307e017a924bf3441223e023d1f"}]},"qwen_image_edit_2511_fp8_lightning":{"description":"Qwen Image Edit 2511 Lightning 4-step speed LoRA","files":[{"tensorFile":"Qwen-Image-Edit-2511-Lightning-4steps-V1.0-bf16.safetensors","description":"Qwen Image Edit 2511 Lightning 4-Step LoRA","destinationFolder":"models/loras","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/loras/Qwen-Image-Edit-2511-Lightning-4steps-V1.0-bf16.safetensors","bytes":849608296,"bytesExact":true,"sha256":"22226e8d05d354bb356627d428809f5afd7819399b077238a2b70a82883a904f"}]},"flux1_shared":{"description":"Models shared by all FLUX.1 workflows (VAE, text encoders)","files":[{"tensorFile":"ae.safetensors","description":"FLUX.1 VAE (shared with Z-Image Turbo)","destinationFolder":"models/vae","tensorDownload":"https://cdn.sogni.ai/vae/ae.safetensors","bytes":335304388,"sha256":"afc8e28272cd15db3919bacdb6918ce9c1ed22e96cb12c4d5ed0fba823529e38"},{"tensorFile":"clip_l.safetensors","description":"FLUX.1 CLIP-L Text Encoder","destinationFolder":"models/text_encoders","tensorDownload":"https://cdn.sogni.ai/text_encoder/clip_l.safetensors","bytes":246144152,"sha256":"660c6f5b1abae9dc498ac2d21e1347d2abdb0cf6c0c0c8576cd796491d9a6cdd"},{"tensorFile":"t5xxl_fp8_e4m3fn_scaled.safetensors","description":"FLUX.1 T5-XXL Text Encoder (FP8)","destinationFolder":"models/text_encoders","tensorDownload":"https://cdn.sogni.ai/text_encoder/t5xxl_fp8_e4m3fn_scaled.safetensors","bytes":5157348688,"bytesExact":true,"sha256":"a498f0485dc9536735258018417c3fd7758dc3bccc0a645feaa472b34955557a"}]},"flux1-schnell-fp8":{"description":"FLUX.1 [schnell] - Fast 4-step text-to-image model","files":[{"tensorFile":"flux1-schnell_fp8.safetensors","description":"FLUX.1 Schnell Diffusion Model (FP8)","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/unet/flux1-schnell_fp8.safetensors","bytes":11891286928,"bytesExact":true,"sha256":"ece1aec579525fc0dc23a6dd9ad0005cefdec58e02d9d1985a78b53df2a48652"}]},"chroma-v.46-flash_fp8":{"description":"Chroma v.46 [flash] - Fast 10-step high-quality text-to-image model","files":[{"tensorFile":"chroma-unlocked-v46-flash_float8_e4m3fn_scaled_learned.safetensors","description":"Chroma v.46 Flash Diffusion Model (FP8)","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/unet/chroma-unlocked-v46-flash_float8_e4m3fn_scaled_learned.safetensors","bytes":8902171990,"bytesExact":true,"sha256":"e18c3ed3ebfc2bf97b29c4d96e6c4c56aa22416f244431fda0d83c8721057dc7"}]},"chroma-v48-detail-svd_fp8":{"description":"Chroma v.48 [detail] - High-detail 20-step text-to-image model","files":[{"tensorFile":"chroma-unlocked-v48-detail-calibrated_float8_e4m3fn_scaled_learned_svd.safetensors","description":"Chroma v.48 Detail Diffusion Model (FP8)","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/unet/chroma-unlocked-v48-detail-calibrated_float8_e4m3fn_scaled_learned_svd.safetensors","bytes":8902171990,"bytesExact":true,"sha256":"081baf829cdb6fe3e614374e278c584e11c295e6c6c3a9d5a0321fbf9ff5e28b"}]},"chroma1-hd_fp8_scaled":{"description":"Chroma1-HD - Final high-resolution Chroma text-to-image model","files":[{"tensorFile":"Chroma1-HD-fp8_scaled_defaultloader_hybrid_large_rev2.safetensors","description":"Chroma1-HD Diffusion Model (FP8 scaled default-loader hybrid large rev2)","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/diffusion_models/Chroma1-HD-fp8_scaled_defaultloader_hybrid_large_rev2.safetensors","bytes":9193371409,"sha256":"f8df6efac8f9c4e778ec07b9c9362d43612cd6874271d602df90b34f72552931"}]},"ltx23-22b-fp8_shared":{"description":"Models shared by all LTX-2.3 22B workflows","files":[{"tensorFile":"gemma_3_12B_it_fp4_mixed.safetensors","description":"LTX-2.3 Text Encoder (Gemma 3 12B FP4)","destinationFolder":"models/text_encoders","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/text_encoders/gemma_3_12B_it_fp4_mixed.safetensors","bytes":9447702218,"bytesExact":true,"sha256":"aaca463d11e6d8d2a4bdb0d6299214c15ef78a3f73e0ef8113d5a9d0219b3f6d"},{"tensorFile":"ltx-2.3_text_projection_bf16.safetensors","description":"LTX-2.3 Text Projection Model","destinationFolder":"models/text_encoders","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/text_encoders/ltx-2.3_text_projection_bf16.safetensors","bytes":2312149072,"bytesExact":true,"sha256":"911d59bb4cb7708179c9a0045ea0fe41212ecfb77aed3a02702b7c0a8274911f"},{"tensorFile":"pruna_ltx2.3_vae_comfy_bf16.safetensors","description":"LTX-2.3 PrunaVAED video decoder","destinationFolder":"models/vae","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/vae/pruna_ltx2.3_vae_comfy_bf16.safetensors","bytes":1327910922,"sha256":"04690af9832b1fa1ec9bd2118e5870b858fbc526998b41cfec9b0017d309edfc"},{"tensorFile":"LTX23_audio_vae_bf16.safetensors","description":"LTX-2.3 Audio VAE","destinationFolder":"models/checkpoints","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/checkpoints/LTX23_audio_vae_bf16.safetensors","bytes":364855188,"bytesExact":true,"sha256":"5bc10fa4adecf99dda132d916e23048cbd56797702c5fa50eb5d2079048a38c3"}]},"ltx23-22b-fp8":{"description":"LTX-2.3 22B distilled diffusion model","files":[{"tensorFile":"ltx-2.3-22b-distilled_transformer_only_fp8_scaled.safetensors","description":"LTX-2.3 22B Distilled Diffusion Model (FP8)","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/diffusion_models/ltx-2.3-22b-distilled_transformer_only_fp8_scaled.safetensors","bytes":23470368784,"bytesExact":true,"sha256":"412ad3606a480963666ad44ad9e0edea5360ee304ff2395cf9e0017e8cc53474"}]},"ltx23-22b-fp8_dev":{"description":"LTX-2.3 22B dev (non-distilled) diffusion model","files":[{"tensorFile":"ltx-2.3-22b-dev_transformer_only_fp8_scaled.safetensors","description":"LTX-2.3 22B Dev Diffusion Model (FP8)","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/diffusion_models/ltx-2.3-22b-dev_transformer_only_fp8_scaled.safetensors","bytes":23470368720,"bytesExact":true,"sha256":"f0b8f92f8a33daf7e7f508878db6d6f25f9c8bc7b90fcb5650fb402a9d9e9f13"}]},"ltx23-22b-fp8_distilled_lora":{"description":"LTX-2.3 22B Distilled LoRA for Stage 2 refinement","files":[{"tensorFile":"ltx-2.3-22b-distilled-lora-dynamic_fro09_avg_rank_105_bf16.safetensors","description":"LTX-2.3 Distilled LoRA (compressed rank 105) - Required for dev model Stage 2 upscale refinement","destinationFolder":"models/loras","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/loras/ltx-2.3-22b-distilled-lora-dynamic_fro09_avg_rank_105_bf16.safetensors","bytes":2586318182,"bytesExact":true,"sha256":"289441f530ca30520bd5e1d4f34b8437f75d6c66f9b40e32b3824e225dd96325"}]},"ltx23-22b-fp8_upscaler":{"description":"LTX-2.3 Spatial Upscaler (2x)","files":[{"tensorFile":"ltx-2.3-spatial-upscaler-x2-1.1.safetensors","description":"LTX-2.3 Spatial Upscaler (2x) v1.1 - For 2-stage generation pipeline","destinationFolder":"models/latent_upscale_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/latent_upscale_models/ltx-2.3-spatial-upscaler-x2-1.1.safetensors","bytes":995743560,"bytesExact":true,"sha256":"5f416311fa8172b65af67530758964708d29a317b830d689a51143b7f91913ed"}]},"ltx23-22b-fp8_text-projection":{"description":"LTX-2.3 text projection (embeddings connector) as a standalone group","files":[{"tensorFile":"ltx-2.3_text_projection_bf16.safetensors","description":"LTX-2.3 DualLinearProjection text projection - embeddings connector for LTXAVTextEncoderLoader (same file as in ltx23-22b-fp8_shared)","destinationFolder":"models/text_encoders","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/text_encoders/ltx-2.3_text_projection_bf16.safetensors","bytes":2312149072,"bytesExact":true,"sha256":"911d59bb4cb7708179c9a0045ea0fe41212ecfb77aed3a02702b7c0a8274911f"}]},"ltx23-22b-fp8mixed_10eros":{"description":"LTX-2.3 10Eros v1.4 uncensored i2v finetune (full fp8mixed checkpoint, abliterated TE, DMD LoRA)","minVramGB":30,"spicy":true,"files":[{"tensorFile":"10Eros_v1.4_fp8mixed_learned.safetensors","description":"10Eros v1.4 full checkpoint (fp8mixed learned) - DiT + video/audio VAEs + text projection","destinationFolder":"models/checkpoints","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/checkpoints/10Eros_v1.4_fp8mixed_learned.safetensors","bytes":29161843630,"sha256":"54bcb40427ff1a3e54cfa6087df765b617bee238d58aa49947c517a727722ad2"},{"tensorFile":"gemma-3-12b-it-ablit-norms-biproj-fp8mixed.safetensors","description":"Abliterated Gemma-3-12B text encoder (fp8mixed) - required for uncensored prompt adherence","destinationFolder":"models/text_encoders","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/text_encoders/gemma-3-12b-it-ablit-norms-biproj-fp8mixed.safetensors","bytes":12778566778,"sha256":"0d76ceae7c1e8cb87dc88c29f37f171819e3bda0d605f196cd5daa1d3cdfd008"},{"tensorFile":"LTX2.3_DMD_reshaped_r256.safetensors","description":"LTX-2.3 DMD distillation LoRA (reshaped r256) - baked at 1.0 in the 10Eros templates","destinationFolder":"models/loras","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/loras/LTX2.3_DMD_reshaped_r256.safetensors","bytes":5095405082,"sha256":"4543ac32e5a3d1b28dac62a956987cced24311a54710d6e175f6fd607a59c46d"},{"tensorFile":"pruna_ltx2.3_vae_comfy_bf16.safetensors","description":"LTX-2.3 PrunaVAED video decoder","destinationFolder":"models/vae","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/vae/pruna_ltx2.3_vae_comfy_bf16.safetensors","bytes":1327910922,"sha256":"04690af9832b1fa1ec9bd2118e5870b858fbc526998b41cfec9b0017d309edfc"}]},"ltx23-22b-fp8_id-lora-talkvid-3k":{"description":"LTX-2.3 22B ID-LoRA TalkVid-3K for speaker identity transfer","files":[{"tensorFile":"LTX-2.3-ID-LoRA-TalkVid-3K.safetensors","description":"ID-LoRA TalkVid-3K - speaker voice identity transfer (~5s reference audio)","destinationFolder":"models/loras","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/loras/LTX-2.3-ID-LoRA-TalkVid-3K.safetensors","bytes":1160000000,"sha256":"e5af73441743b4852f228b03e444888dff3da80d2666033af2367ab7bda6d8b9"}]},"ltx23-22b-fp8_ic-lora-union-control":{"description":"LTX-2.3 22B IC-LoRA Union Control (canny+depth+pose in single LoRA)","files":[{"tensorFile":"ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","description":"Union Control IC-LoRA with half-resolution reference for v2v workflows","destinationFolder":"models/loras","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/loras/ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors","bytes":654000000,"sha256":"149e34b148e15f6fea8b0a86e949264f6ed47183a073b30683b5e851a6b9d3f1"}]},"ltx23-22b-fp8_ic-lora-in-outpaint":{"description":"LTX-2.3 22B IC-LoRA In/Outpainting (official) for video inpaint + canvas expansion","files":[{"tensorFile":"ltx-2.3-22b-ic-lora-in-outpainting-0.9.safetensors","description":"Official In/Outpainting IC-LoRA (0.9) for v2v inpaint and outpaint workflows","destinationFolder":"models/loras","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/loras/ltx-2.3-22b-ic-lora-in-outpainting-0.9.safetensors","bytes":1308778338,"sha256":"73dd0841c0d4f0eb26fb1f017781b841b2752021944ac5ecefe57917f6dae6b5"}]},"ltx25-shared":{"description":"Official shared assets for all LTX-2.5 22B workflows","files":[{"tensorFile":"gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","description":"LTX-2.5 Gemma 4 12B text encoder with projection (Comfy INT8-ConvRot)","destinationFolder":"models/text_encoders","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/text_encoders/gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors","bytes":15372971786,"bytesExact":true,"sha256":"09a89e084de1a149c3de60cfe9dfd3e5161967eb09eea39e806fcdeffdd568de"},{"tensorFile":"ltx-2.5-video-vae-bf16.safetensors","description":"LTX-2.5 Video VAE (BF16)","destinationFolder":"models/vae","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/vae/ltx-2.5-video-vae-bf16.safetensors","bytes":1472223346,"bytesExact":true,"sha256":"847e14ca7f3355debca0cea4eaa24ac0fbcdf0061da054ac89ca638a869ddba3"},{"tensorFile":"ltx-2.5-audio-vae-bf16.safetensors","description":"LTX-2.5 Audio VAE (BF16)","destinationFolder":"models/vae","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/vae/ltx-2.5-audio-vae-bf16.safetensors","bytes":364866540,"bytesExact":true,"sha256":"c52733d37f6a7fb7949c3dc0fb468c6cb2169e4d836983a73babb9f0d54837a5"}]},"ltx25-distilled":{"description":"Official LTX-2.5 22B distilled transformer (Comfy INT8-ConvRot)","files":[{"tensorFile":"ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","description":"LTX-2.5 22B Distilled Transformer (Comfy INT8-ConvRot)","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/diffusion_models/ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors","bytes":21504034224,"bytesExact":true,"sha256":"c4279eeff115cbeaca494bd2183e7d768c38fe85a184dc6afbb7159157c44334"}]},"ltx25-upscaler":{"description":"Official LTX-2.5 spatial latent upscaler for two-stage workflows","files":[{"tensorFile":"ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","description":"LTX-2.5 Spatial Latent Upscaler x2 v1.0 (BF16)","destinationFolder":"models/latent_upscale_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/latent_upscale_models/ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors","bytes":995778752,"bytesExact":true,"sha256":"eb5a71fe4068ee87ccdb1c3aa635e547ca76bd2d30ae20ae889f2c325c0677e8"}]},"qwen_image_2512_fp8_shared":{"description":"Models shared by all Qwen Image 2512 workflows","files":[{"tensorFile":"qwen_2.5_vl_7b_fp8_scaled.safetensors","description":"Qwen 2.5 VL Text Encoder (FP8)","destinationFolder":"models/text_encoders","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/text_encoders/qwen_2.5_vl_7b_fp8_scaled.safetensors","bytes":9384670680,"sha256":"cb5636d852a0ea6a9075ab1bef496c0db7aef13c02350571e388aea959c5c0b4"},{"tensorFile":"qwen_image_vae.safetensors","description":"Qwen Image VAE","destinationFolder":"models/vae","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/vae/qwen_image_vae.safetensors","bytes":253806246,"sha256":"a70580f0213e67967ee9c95f05bb400e8fb08307e017a924bf3441223e023d1f"}]},"qwen_image_2512_fp8":{"description":"Qwen Image 2512 diffusion model (FP8)","files":[{"tensorFile":"qwen_image_2512_fp8_e4m3fn.safetensors","description":"Qwen Image 2512 Diffusion Model (FP8)","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/diffusion_models/qwen_image_2512_fp8_e4m3fn.safetensors","bytes":20430679144,"bytesExact":true,"sha256":"5dc80554d5d83390046a2f4a94ece06afb7700bf7b0aaf8bde9769793875876b"}]},"qwen_image_2512_lightning":{"description":"Qwen Image Lightning LoRA for fast 4-step generation","files":[{"tensorFile":"Qwen-Image-Lightning-4steps-V1.0.safetensors","description":"Qwen Image Lightning LoRA (4-step)","destinationFolder":"models/loras","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/loras/Qwen-Image-Lightning-4steps-V1.0.safetensors","bytes":1698951104,"bytesExact":true,"sha256":"9526e90d71c4290392feeccf3c2172cb77ab3a489f1faeb956637f97acb4c8b1"}]},"z_image_bf16_shared":{"description":"Models shared by Z-Image turbo and non-turbo (text encoder + VAE)","files":[{"tensorFile":"qwen_3_4b.safetensors","description":"Z-Image Text Encoder (Qwen 3 4B)","destinationFolder":"models/text_encoders","tensorDownload":"https://pub-5bc58981af9f42659ff8ada57bfea92c.r2.dev/ComfyUI/models/text_encoders/qwen_3_4b.safetensors","bytes":8044982048,"sha256":"6c671498573ac2f7a5501502ccce8d2b08ea6ca2f661c458e708f36b36edfc5a"},{"tensorFile":"ae.safetensors","description":"Z-Image VAE","destinationFolder":"models/vae","tensorDownload":"https://pub-5bc58981af9f42659ff8ada57bfea92c.r2.dev/vae/ae.safetensors","bytes":335304388,"sha256":"afc8e28272cd15db3919bacdb6918ce9c1ed22e96cb12c4d5ed0fba823529e38"}]},"z_image_bf16":{"description":"Z-Image text-to-image model (non-turbo, higher quality)","files":[{"tensorFile":"z_image_bf16.safetensors","description":"Z-Image Diffusion Model","destinationFolder":"models/diffusion_models","tensorDownload":"https://pub-5bc58981af9f42659ff8ada57bfea92c.r2.dev/ComfyUI/models/diffusion_models/z_image_bf16.safetensors","bytes":13206016000,"sha256":"996a67d3ff666946b1c25cbc16d1b1918b6cc0ac166309e23fe3b3d830263dee"}]},"ace_step_1.5_shared":{"description":"ACE-Step 1.5 shared models (0.6B caption encoder + 4B LM + VAE) - used by both SFT and Turbo","files":[{"tensorFile":"qwen_0.6b_ace15.safetensors","description":"ACE-Step 1.5 Caption Encoder (0.6B - Qwen3, DualCLIPLoader clip_name1)","destinationFolder":"models/clip","tensorDownload":"https://pub-5bc58981af9f42659ff8ada57bfea92c.r2.dev/ComfyUI/models/clip/qwen_0.6b_ace15.safetensors","bytes":1191588248,"bytesExact":true,"sha256":"fd4590c82153b8ddb67e15a2e7aaa8afa8b83a858c8a9b82a4831063156aa7a7"},{"tensorFile":"qwen_4b_ace15.safetensors","description":"ACE-Step 1.5 Language Model (4B - Qwen3, DualCLIPLoader clip_name2)","destinationFolder":"models/clip","tensorDownload":"https://pub-5bc58981af9f42659ff8ada57bfea92c.r2.dev/ComfyUI/models/clip/qwen_4b_ace15.safetensors","bytes":8379154232,"bytesExact":true,"sha256":"ffe5ffb855086c2ab55e467e9859fb01894781020a0376484dd19de166b79873"},{"tensorFile":"ace_1.5_vae.safetensors","description":"ACE-Step 1.5 VAE (Audio Decoder)","destinationFolder":"models/vae","tensorDownload":"https://pub-5bc58981af9f42659ff8ada57bfea92c.r2.dev/ComfyUI/models/vae/ace_1.5_vae.safetensors","bytes":337431732,"bytesExact":true,"sha256":"6de92e3a862acd287e08b024ac90f0783a8635451b728721a33ff03565bcb2bb"}]},"ace_step_1.5_sft":{"description":"ACE-Step 1.5 SFT DiT - high-quality music generation with CFG guidance","files":[{"tensorFile":"acestep_v1.5_sft.safetensors","description":"ACE-Step 1.5 SFT DiT (Diffusion Model)","destinationFolder":"models/diffusion_models","tensorDownload":"https://pub-5bc58981af9f42659ff8ada57bfea92c.r2.dev/ComfyUI/models/diffusion_models/acestep_v1.5_sft.safetensors","bytes":4787825604,"sha256":"d4dd3a93870f06720027965b90771f529ab02094b3d29e2518f1d5e097e1af7e"}]},"ace_step_1.5_turbo":{"description":"ACE-Step 1.5 Turbo DiT - fast music generation in 4-16 steps, no CFG support","files":[{"tensorFile":"acestep_v1.5_turbo.safetensors","description":"ACE-Step 1.5 Turbo DiT (Diffusion Model)","destinationFolder":"models/diffusion_models","tensorDownload":"https://pub-5bc58981af9f42659ff8ada57bfea92c.r2.dev/ComfyUI/models/diffusion_models/acestep_v1.5_turbo.safetensors","bytes":4787825604,"bytesExact":true,"sha256":"3f6e0797fad420a39bd33979eb6e840e30989e34a3794e843d23b60ec6e422d7"}]},"ace_step_1.5_xl_sft":{"description":"ACE-Step 1.5 XL SFT DiT - 4B diffusion model with CFG guidance","minVramGB":20,"files":[{"tensorFile":"acestep_v1.5_xl_sft_bf16.safetensors","description":"ACE-Step 1.5 XL SFT DiT (BF16 Diffusion Model)","destinationFolder":"models/diffusion_models","tensorDownload":"https://pub-5bc58981af9f42659ff8ada57bfea92c.r2.dev/ComfyUI/models/diffusion_models/acestep_v1.5_xl_sft_bf16.safetensors","bytes":9974719930,"sha256":"3c05ae268353b3540fb1fd7db4fd77ffbda9802ec641b624e15648e030ecf3ce"}]},"ace_step_1.5_xl_turbo":{"description":"ACE-Step 1.5 XL Turbo DiT - 4B diffusion model, fast 8-step generation, no CFG support","minVramGB":20,"files":[{"tensorFile":"acestep_v1.5_xl_turbo_bf16.safetensors","description":"ACE-Step 1.5 XL Turbo DiT (BF16 Diffusion Model)","destinationFolder":"models/diffusion_models","tensorDownload":"https://pub-5bc58981af9f42659ff8ada57bfea92c.r2.dev/ComfyUI/models/diffusion_models/acestep_v1.5_xl_turbo_bf16.safetensors","bytes":9974719892,"sha256":"86a1afb0a1f711f0e3304ff65d874df3ae6783db683dcf982513fb9b6d14ae71"}]},"dark_beast_z_image_turbo_v9_bf16":{"description":"Dark Beast Z-Image Turbo v9 community fine-tune for Z-Image Turbo (CivitAI version DBZiT9 DIM RClaw, BF16)","minVramGB":16,"spicy":true,"files":[{"tensorFile":"dark_beast_z_image_turbo_bf16.safetensors","description":"Dark Beast Z-Image Turbo v9 DBZiT9 DIM RClaw Diffusion Model (BF16)","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/diffusion_models/dark_beast_z_image_turbo_bf16.safetensors","bytes":12309878456,"sha256":"bf2937ae4edc0183ebfe180834208ea7ac0624cd933aafa6cda05a21fdb8e927"}]},"dark_beast_krea2_fp8":{"description":"Dark Beast Krea 2 v3 INT8 ConvRot community fine-tune for Krea 2 (13.16GB CivitAI diffusion-model file, model 2242173 version 3173268). Replaces the v1 FP8 file, which was a flat 8-bit cast with no scales.","minVramGB":16,"spicy":true,"files":[{"tensorFile":"dark_beast_krea2_v3.0_int8_convrot.safetensors","description":"Dark Beast Krea 2 v3 INT8 ConvRot Diffusion Model","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/diffusion_models/dark_beast_krea2_v3.0_int8_convrot.safetensors","bytes":14132235920,"sha256":"b60cb86fc1c8a84f37991c0c4d9bffe9ba4a2ce6ae1ec26ff2691bb21d87c433"},{"tensorFile":"qwen3vl_4b_fp8_scaled.safetensors","description":"Krea 2 Text Encoder (Qwen3-VL-4B FP8 scaled)","destinationFolder":"models/text_encoders","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/text_encoders/qwen3vl_4b_fp8_scaled.safetensors","bytes":5242467968,"sha256":"54bd5144df0bbc25dd6ccadfcb826b521445a1b06ae5a42570bdd2974ca87094"},{"tensorFile":"krea2RealVae_v10.safetensors","description":"Krea 2 Real VAE v1.0","destinationFolder":"models/vae","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/vae/krea2RealVae_v10.safetensors","bytes":507591212,"sha256":"0dbbe0baeca04c2b98d2f3809c6f595608939809c88b695ba971368f17c874b8"},{"tensorFile":"qwen_image_vae.safetensors","description":"Qwen Image VAE (native Krea 2 option)","destinationFolder":"models/vae","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/vae/qwen_image_vae.safetensors","bytes":253806246,"sha256":"a70580f0213e67967ee9c95f05bb400e8fb08307e017a924bf3441223e023d1f"},{"tensorFile":"wan_2.1_vae.safetensors","description":"WAN 2.1 VAE (popular alternate Krea 2 VAE)","destinationFolder":"models/vae","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/vae/wan_2.1_vae.safetensors","bytes":253815318,"sha256":"2fc39d31359a4b0a64f55876d8ff7fa8d780956ae2cb13463b0223e15148976b"}]},"one_obsession_v22_fp16":{"description":"One Obsession v22 Illustrious XL-based anime checkpoint (CivitAI v22, FP16)","files":[{"tensorFile":"oneObsession_v22.safetensors","description":"One Obsession v22 Illustrious Checkpoint (FP16)","destinationFolder":"models/checkpoints","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/checkpoints/oneObsession_v22.safetensors","bytes":6938041144,"sha256":"42cba375d474febcdbea5422624121c295142ac1b0feaa153271ca2cd387ed10"}]},"minimax-h3-shared":{"description":"MiniMax H3 shared encoder and native video/audio decoders, used by every H3 workflow (FL2VA and Ref2VA), plus the 0.69 GB latent upscaler that serves the two-stage FastH3 model ids (worker 1.0.217+ receives them as the base FastH3 workflow with outputScale: 2).","minVramGB":23,"files":[{"tensorFile":"qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors","description":"MiniMax H3 Qwen3-VL 32B NVFP4 AWQ text encoder","destinationFolder":"models/text_encoders","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors","bytes":15687142551,"bytesExact":true,"sha256":"35a88d51044231fe332301d7a62aa81e3f2cba62febeb446e2c1e3e0ef76f2c6","required":true},{"tensorFile":"minimax_h3_video_vae_fp16.safetensors","description":"MiniMax H3 video VAE","destinationFolder":"models/vae","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/vae/minimax_h3_video_vae_fp16.safetensors","bytes":5207808496,"bytesExact":true,"sha256":"7c1f131492e7eddacaac9069a61b81bdd39de5cc96561e677c5eab1cdce5e522","required":true},{"tensorFile":"minimax_h3_audio_vae_fp32.safetensors","description":"MiniMax H3 native stereo audio VAE","destinationFolder":"models/vae","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/vae/minimax_h3_audio_vae_fp32.safetensors","bytes":605254808,"bytesExact":true,"sha256":"8e505d95dd1561d47abd43d4238fd40d9bb1ae9e147ed0a4cba778d76ae4db48","required":true},{"tensorFile":"minimax_h3_latent_upscaler_3d_fp16.safetensors","description":"MiniMax H3 2K learned latent upscaler (Comfyui_Minimax_h3_latent_Upscaler by LBH-123-AI; Apache-2.0 weights from Hugging Face LBH-123-AI/Minimax_h3_latent_Upscaler @ 13ccf95d). It serves the two-stage FastH3 model ids (minimax-h3-fastvideo-int8_{t2v,i2v,flf2v,ia2v,flfa2v,a2v}_turbo_2stage): the socket sends worker 1.0.217+ (per-step averaged 2K refinement tiles; 1.0.215 and 1.0.216 no longer receive two-stage jobs) the base FastH3 workflow with outputScale: 2, which enlarges the video latent 2x and refines it so the clip delivers twice its canvas (1344x768 -> 2688x1536); inert on older workers","destinationFolder":"models/latent_upscale_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/latent_upscale_models/minimax_h3_latent_upscaler_3d_fp16.safetensors","bytes":690592672,"bytesExact":true,"sha256":"043e5a48e161610ef6c3ea974645220354d06fa618abca15f76d084812eb55c2","required":true}]},"minimax-h3-fl2va-fp8":{"description":"MiniMax H3 FL2VA pruned FP8-scaled diffusion model for text, first-frame, and first/last-frame video with audio.","minVramGB":32,"files":[{"tensorFile":"minimax_h3_fl2va_pruned_fp8_scaled.safetensors","description":"MiniMax H3 FL2VA pruned FP8-scaled model","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/diffusion_models/minimax_h3_fl2va_pruned_fp8_scaled.safetensors","bytes":20958205608,"bytesExact":true,"sha256":"12944c1f7791637e7de12208aef04da82bd26b95271b1b47d817364315ade993","required":true}]},"minimax-h3-fastvideo-int8":{"description":"FastVideo VSA data-free 4-step MiniMax H3 FL2VA distillation in Kijai's INT8 ConvRot ComfyUI format. Used by worker 1.0.193+ FastH3 t2v/i2v/flf2v graphs and the worker 1.0.217+ FastH3 audio-guided graphs (image + audio ia2v, first/last frame + audio flfa2v, audio-only a2v) with the validated learned-gate VSA recipe; reuses the shared H3 encoder and VAEs.","minVramGB":23,"files":[{"tensorFile":"minimax_h3_fastvideo_vsa_datafree_1300step_4step_int8_convrot.safetensors","description":"MiniMax H3 FastVideo VSA data-free 4-step INT8 ConvRot checkpoint","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/diffusion_models/minimax_h3_fastvideo_vsa_datafree_1300step_4step_int8_convrot.safetensors","bytes":22898594920,"bytesExact":true,"sha256":"7221ae65d78780354d51e5048d29728d9f1f8fb9baf50b1dd3df85f5101413d3","required":true}]},"minimax-h3-ref2va-fp8":{"description":"MiniMax H3 Ref2VA pruned FP8-scaled diffusion model for multi-reference (image/video/audio) video with audio. Separate checkpoint from FL2VA; reuses the minimax-h3-shared encoder and VAEs.","minVramGB":32,"files":[{"tensorFile":"minimax_h3_ref2va_pruned_fp8_scaled.safetensors","description":"MiniMax H3 Ref2VA pruned FP8-scaled model","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/diffusion_models/minimax_h3_ref2va_pruned_fp8_scaled.safetensors","bytes":20958205608,"bytesExact":true,"sha256":"f86f2f79ebd2d76eb8eeb46091e83982e6ff51d255747e7b16e92834b392b8e9","required":true}]},"minimax-h3-fl2va-balanced-lora":{"description":"MiniMax H3 FL2VA Balanced acceleration assets. Worker 1.0.213 graphs (t2v/i2v/flf2v Balanced) apply Larry's v4 step-600 EMA adapter at strength 1.0 through his resident-weight loader with 8-step Euler/simple (ER-SDE selectable) and 6/3 video/audio sigma shifts by default (the video shift is user-selectable 4-12). LightX2V's official v1.0 8-step 768p FL2VA LoRA remains for the 1.0.205-1.0.212 FL2VA Balanced graphs, which name it; drop it once no worker below 1.0.213 is reporting in.","minVramGB":32,"files":[{"tensorFile":"minimax_h3_turbo_v4_step600_ema.safetensors","description":"Larry MiniMax H3 Turbo v4 step-600 EMA community adapter (Apache-2.0), the 8-step Balanced LoRA for every H3 mode from worker 1.0.213; one file covers the FL2VA and REF2VA bases","destinationFolder":"models/loras","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/loras/minimax_h3_turbo_v4_step600_ema.safetensors","bytes":779849816,"bytesExact":true,"sha256":"5f3a626cd72c93a8b9318d6760c510bc5092d2ab13aaba1f932c5bab07a416d3","required":true},{"tensorFile":"minimax_h3_fl2v_turbo_8step_v1.0_768p_comfyui_bf16.safetensors","description":"Superseded on FL2VA Balanced: official LightX2V MiniMax H3 Turbo v1.0 8-step 768p FL2VA LoRA in full-rank ComfyUI BF16 format, which worker 1.0.205-1.0.212 graphs name; retained for those graphs during migration","destinationFolder":"models/loras","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/loras/minimax_h3_fl2v_turbo_8step_v1.0_768p_comfyui_bf16.safetensors","bytes":1956193000,"bytesExact":true,"sha256":"08cfe946033af7d27719b964b6e0a0e50c32138daabbd6ce4137e23df6bf9980","required":true}]},"minimax-h3-ref2va-balanced-lora":{"description":"MiniMax H3 Ref2VA Balanced acceleration assets. Worker 1.0.213 graphs apply Larry's v4 step-600 EMA adapter at strength 1.0 through his resident-weight loader with 8-step Euler/simple (ER-SDE selectable) and 6/3 video/audio sigma shifts by default (1.0.200-1.0.212 used H3's native 12/3; the video shift is user-selectable 4-12).","minVramGB":32,"files":[{"tensorFile":"minimax_h3_turbo_v4_step600_ema.safetensors","description":"Larry MiniMax H3 Turbo v4 step-600 EMA community adapter (Apache-2.0), the 8-step Balanced LoRA for every H3 mode from worker 1.0.200; one file covers the FL2VA and REF2VA bases","destinationFolder":"models/loras","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/loras/minimax_h3_turbo_v4_step600_ema.safetensors","bytes":779849816,"bytesExact":true,"sha256":"5f3a626cd72c93a8b9318d6760c510bc5092d2ab13aaba1f932c5bab07a416d3","required":true}]},"minimax-h3-fl2va-turbo-lora":{"description":"MiniMax H3 Turbo LoRAs by LightX2V (Apache-2.0), 4-step 768p full-rank ComfyUI BF16 releases distilled at the native 1344x768 Turbo output size and used at strength 1.0 with 6/3 video/audio sigma shifts. Worker 1.0.208 and later graphs use v1.0, restored after native-frame comparisons found worse motion artifacts with v1.1. The v1.1 asset remains for the 1.0.205-1.0.207 graphs, which name it; drop it once no worker below 1.0.208 is reporting in.","minVramGB":32,"files":[{"tensorFile":"minimax_h3_fl2v_turbo_4step_v1.1_768p_comfyui_bf16.safetensors","description":"Superseded official LightX2V MiniMax H3 Turbo v1.1 4-step 768p FL2V LoRA in full-rank ComfyUI BF16 format, retained for worker 1.0.205-1.0.207 Turbo graphs during migration","destinationFolder":"models/loras","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/loras/minimax_h3_fl2v_turbo_4step_v1.1_768p_comfyui_bf16.safetensors","bytes":1956192992,"bytesExact":true,"sha256":"449d80f301ac571622c72e28b8fd72a4b3681b7a8df8a92f17c8f6ec43f56558","required":true},{"tensorFile":"minimax_h3_fl2v_turbo_4step_v1.0_768p_comfyui_bf16.safetensors","description":"Official LightX2V MiniMax H3 Turbo v1.0 4-step 768p FL2V LoRA in full-rank ComfyUI BF16 format, distilled at native 1344x768; the in-service Turbo LoRA for worker 1.0.208 and later graphs","destinationFolder":"models/loras","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/loras/minimax_h3_fl2v_turbo_4step_v1.0_768p_comfyui_bf16.safetensors","bytes":1956192992,"bytesExact":true,"sha256":"c396a9a06f58399e9df9754b18299818d84a2ddd371724ba48fe4a41221437dc","required":true}]},"krea2_identity_edit_sogni_v0_3_alpha":{"description":"Sogni Krea 2 Identity Edit v0.3 Alpha - Sogni-trained identity-preserving image editing for Krea 2. Fine-tuned by Sogni on top of the community krea2-identity-edit v1.2 (conradlocke) with calibrated reframing, directional gaze control, subject edits, and tiered photo restoration. v0.3 adds a 1024px finishing pass and materially improves full-color restoration while preserving identity.","minVramGB":16,"files":[{"tensorFile":"krea2_identity_edit_sogni_v0_3_alpha.safetensors","description":"Sogni Krea 2 Identity Edit v0.3 Alpha LoRA (rank 64), trained by Sogni","destinationFolder":"models/loras","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/loras/krea2_identity_edit_sogni_v0_3_alpha.safetensors","bytes":457111488,"sha256":"e7202aca20ccce17c6302449a5981bc1dbaa3fa45e0ba6801e21641ee06e0135"}]},"minimax_music3":{"description":"MiniMax Music 3 - hierarchical AR music model (8B planner + RVQ depth decoder + 2.4B flow-matching DiT + DAV decoder)","minVramGB":24,"files":[{"tensorFile":"minimax_music3_dit_fp16.safetensors","description":"MiniMax Music 3 DiT (FP16 Diffusion Model)","destinationFolder":"models/diffusion_models","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/diffusion_models/minimax_music3_dit_fp16.safetensors","bytes":4914197682,"sha256":"45494a2b6b69af115902ff28eaf54118d19067aa54da01000f3e3efce7ba0e34"},{"tensorFile":"minimax_music3_text_encoder_pruned_int8_convrot.safetensors","description":"MiniMax Music 3 Text Encoder (Pruned INT8 ConvRot - 8B planner + RVQ depth decoder)","destinationFolder":"models/text_encoders","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/text_encoders/minimax_music3_text_encoder_pruned_int8_convrot.safetensors","bytes":9196611886,"sha256":"010b7416d2336a08c711bc22ee65849c9623069ddb7d89bec011a75699e52014"},{"tensorFile":"minimax_music3_dav.safetensors","description":"MiniMax Music 3 DAV (Flow-VAE Audio Decoder)","destinationFolder":"models/vae","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/vae/minimax_music3_dav.safetensors","bytes":216696128,"sha256":"2a32155b769be01445fcc2a8663b910fc9e1751e18dc1c3ec528064512d9ef0c"}]},"minimax-h3-ref2va-turbo-lora":{"description":"Dedicated LightX2V MiniMax H3 Ref2VA/R2V Turbo v0.1 LoRA (Apache-2.0), used at the official four-step Euler/simple design point with strength 1.0 and 12/3 video/audio sigma shifts.","minVramGB":32,"files":[{"tensorFile":"minimax_h3_ref2v_turbo_4step_v0.1_comfyui_bf16.safetensors","description":"Official LightX2V MiniMax H3 Ref2VA Turbo v0.1 four-step LoRA in full-rank ComfyUI BF16 format","destinationFolder":"models/loras","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/loras/minimax_h3_ref2v_turbo_4step_v0.1_comfyui_bf16.safetensors","bytes":1956193000,"bytesExact":true,"sha256":"5b9ab5ade15d0775676d01a907268a69a1468dc6033b3b0d3ded5502f3ebb84c","required":true}]},"qwen3-tts-custom-voice":{"description":"Qwen3-TTS 12Hz 1.7B CustomVoice - nine studio voices with instruction-driven style control","minVramGB":24,"files":[{"tensorFile":"config.json","description":"Qwen3-TTS CustomVoice model config","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-custom-voice","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-custom-voice/config.json","bytes":4908,"bytesExact":true,"sha256":"17a07f527a1c25ea30b4e023a184482a23d3e279d697b1dc81b1bde498d29cf9","required":true},{"tensorFile":"model.safetensors","description":"Qwen3-TTS 12Hz 1.7B CustomVoice BF16 weights","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-custom-voice","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-custom-voice/model.safetensors","bytes":3833402552,"bytesExact":true,"sha256":"38b1d5971bdbd982b561cccec982669a53b0537c3cf5e9bd4778ed07bb2f5137","required":true},{"tensorFile":"generation_config.json","description":"Sampling defaults","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-custom-voice","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-custom-voice/generation_config.json","bytes":245,"bytesExact":true,"sha256":"f1b90b4513f3b34c62851049e2492d7b4c5940daf1276f89c82b8ef04127f3aa","required":true},{"tensorFile":"merges.txt","description":"Text tokenizer merges","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-custom-voice","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-custom-voice/merges.txt","bytes":1671839,"bytesExact":true,"sha256":"599bab54075088774b1733fde865d5bd747cbcc7a547c5bc12610e874e26f5e3","required":true},{"tensorFile":"preprocessor_config.json","description":"Audio preprocessor config","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-custom-voice","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-custom-voice/preprocessor_config.json","bytes":127,"bytesExact":true,"sha256":"efdde1022ea9d76928bf7a9cd53139138f5ba2e466e837f08f6105ab1af1c119","required":true},{"tensorFile":"tokenizer_config.json","description":"Text tokenizer config","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-custom-voice","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-custom-voice/tokenizer_config.json","bytes":7344,"bytesExact":true,"sha256":"dc3c31c3bdaedd5016382bb3cbe07323026775ad51f5a4fb564505992ae4a670","required":true},{"tensorFile":"vocab.json","description":"Text tokenizer vocabulary","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-custom-voice","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-custom-voice/vocab.json","bytes":2776833,"bytesExact":true,"sha256":"ca10d7e9fb3ed18575dd1e277a2579c16d108e32f27439684afa0e10b1440910","required":true},{"tensorFile":"config.json","description":"12Hz speech codec config","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-custom-voice/speech_tokenizer","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-custom-voice/speech_tokenizer/config.json","bytes":2336,"bytesExact":true,"sha256":"ee65bb901c876664ab8707c487157aa1a6ee57c65969b28fb5ec9dc211e68167","required":true},{"tensorFile":"configuration.json","description":"12Hz speech codec configuration","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-custom-voice/speech_tokenizer","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-custom-voice/speech_tokenizer/configuration.json","bytes":76,"bytesExact":true,"sha256":"6bc26d64eb5024b4d1dab5a52371958b429256d6c9d59787f1f5294a54e0cebd","required":true},{"tensorFile":"preprocessor_config.json","description":"12Hz speech codec preprocessor config","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-custom-voice/speech_tokenizer","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-custom-voice/speech_tokenizer/preprocessor_config.json","bytes":234,"bytesExact":true,"sha256":"fcb3805e597e786d4067706e602f6688524640f8d3396790e2e09b5942fcbdfb","required":true},{"tensorFile":"model.safetensors","description":"Qwen3-TTS 12Hz speech codec (encoder + decoder)","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-custom-voice/speech_tokenizer","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-custom-voice/speech_tokenizer/model.safetensors","bytes":682293092,"bytesExact":true,"sha256":"836b7b357f5ea43e889936a3709af68dfe3751881acefe4ecf0dbd30ba571258","required":true},{"tensorFile":"LICENSE","description":"Apache License 2.0 covering the redistributed Qwen3-TTS weights and runtime","destinationFolder":"licenses/qwen3-tts","tensorDownload":"https://cdn.sogni.ai/ComfyUI/licenses/qwen3-tts/LICENSE","bytes":11343,"bytesExact":true,"sha256":"a44a6081c73ad75f0255bb2bb5cab74ef1829565a895a24e53a4f11290ab7655","required":false}]},"qwen3-tts-voice-clone":{"description":"Qwen3-TTS 12Hz 1.7B Base - zero-shot voice cloning from a short reference recording","minVramGB":24,"files":[{"tensorFile":"config.json","description":"Qwen3-TTS Base model config","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-base","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-base/config.json","bytes":4494,"bytesExact":true,"sha256":"b4f01752d15a488abde3e1ab44723ae4f4b9e68a4037257b098b3737893cc1f9","required":true},{"tensorFile":"model.safetensors","description":"Qwen3-TTS 12Hz 1.7B Base BF16 weights","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-base","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-base/model.safetensors","bytes":3857413744,"bytesExact":true,"sha256":"38fc7fc51c5e776e840414b6fd443962e9411b9654888fd7913e4da643cb857c","required":true},{"tensorFile":"generation_config.json","description":"Sampling defaults","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-base","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-base/generation_config.json","bytes":245,"bytesExact":true,"sha256":"f1b90b4513f3b34c62851049e2492d7b4c5940daf1276f89c82b8ef04127f3aa","required":true},{"tensorFile":"merges.txt","description":"Text tokenizer merges","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-base","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-base/merges.txt","bytes":1671839,"bytesExact":true,"sha256":"599bab54075088774b1733fde865d5bd747cbcc7a547c5bc12610e874e26f5e3","required":true},{"tensorFile":"preprocessor_config.json","description":"Audio preprocessor config","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-base","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-base/preprocessor_config.json","bytes":127,"bytesExact":true,"sha256":"efdde1022ea9d76928bf7a9cd53139138f5ba2e466e837f08f6105ab1af1c119","required":true},{"tensorFile":"tokenizer_config.json","description":"Text tokenizer config","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-base","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-base/tokenizer_config.json","bytes":7344,"bytesExact":true,"sha256":"dc3c31c3bdaedd5016382bb3cbe07323026775ad51f5a4fb564505992ae4a670","required":true},{"tensorFile":"vocab.json","description":"Text tokenizer vocabulary","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-base","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-base/vocab.json","bytes":2776833,"bytesExact":true,"sha256":"ca10d7e9fb3ed18575dd1e277a2579c16d108e32f27439684afa0e10b1440910","required":true},{"tensorFile":"config.json","description":"12Hz speech codec config","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-base/speech_tokenizer","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-base/speech_tokenizer/config.json","bytes":2336,"bytesExact":true,"sha256":"ee65bb901c876664ab8707c487157aa1a6ee57c65969b28fb5ec9dc211e68167","required":true},{"tensorFile":"configuration.json","description":"12Hz speech codec configuration","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-base/speech_tokenizer","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-base/speech_tokenizer/configuration.json","bytes":76,"bytesExact":true,"sha256":"6bc26d64eb5024b4d1dab5a52371958b429256d6c9d59787f1f5294a54e0cebd","required":true},{"tensorFile":"preprocessor_config.json","description":"12Hz speech codec preprocessor config","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-base/speech_tokenizer","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-base/speech_tokenizer/preprocessor_config.json","bytes":234,"bytesExact":true,"sha256":"fcb3805e597e786d4067706e602f6688524640f8d3396790e2e09b5942fcbdfb","required":true},{"tensorFile":"model.safetensors","description":"Qwen3-TTS 12Hz speech codec (encoder + decoder)","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-base/speech_tokenizer","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-base/speech_tokenizer/model.safetensors","bytes":682293092,"bytesExact":true,"sha256":"836b7b357f5ea43e889936a3709af68dfe3751881acefe4ecf0dbd30ba571258","required":true}]},"qwen3-tts-voice-design":{"description":"Qwen3-TTS 12Hz 1.7B VoiceDesign - invents a voice from a written description","minVramGB":24,"files":[{"tensorFile":"config.json","description":"Qwen3-TTS VoiceDesign model config","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-voice-design","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-voice-design/config.json","bytes":4421,"bytesExact":true,"sha256":"aecd2cc4c1fe9edef1cb7ca7c401685a43879ad43f3f9e883f1c6760b61731e0","required":true},{"tensorFile":"model.safetensors","description":"Qwen3-TTS 12Hz 1.7B VoiceDesign BF16 weights","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-voice-design","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-voice-design/model.safetensors","bytes":3833402552,"bytesExact":true,"sha256":"391e8db219f292c515297cdceeb43e4eae67cdde35fa57e79a6a8a532fca0522","required":true},{"tensorFile":"generation_config.json","description":"Sampling defaults","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-voice-design","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-voice-design/generation_config.json","bytes":245,"bytesExact":true,"sha256":"f1b90b4513f3b34c62851049e2492d7b4c5940daf1276f89c82b8ef04127f3aa","required":true},{"tensorFile":"merges.txt","description":"Text tokenizer merges","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-voice-design","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-voice-design/merges.txt","bytes":1671839,"bytesExact":true,"sha256":"599bab54075088774b1733fde865d5bd747cbcc7a547c5bc12610e874e26f5e3","required":true},{"tensorFile":"preprocessor_config.json","description":"Audio preprocessor config","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-voice-design","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-voice-design/preprocessor_config.json","bytes":127,"bytesExact":true,"sha256":"efdde1022ea9d76928bf7a9cd53139138f5ba2e466e837f08f6105ab1af1c119","required":true},{"tensorFile":"tokenizer_config.json","description":"Text tokenizer config","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-voice-design","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-voice-design/tokenizer_config.json","bytes":7344,"bytesExact":true,"sha256":"dc3c31c3bdaedd5016382bb3cbe07323026775ad51f5a4fb564505992ae4a670","required":true},{"tensorFile":"vocab.json","description":"Text tokenizer vocabulary","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-voice-design","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-voice-design/vocab.json","bytes":2776833,"bytesExact":true,"sha256":"ca10d7e9fb3ed18575dd1e277a2579c16d108e32f27439684afa0e10b1440910","required":true},{"tensorFile":"config.json","description":"12Hz speech codec config","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-voice-design/speech_tokenizer","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-voice-design/speech_tokenizer/config.json","bytes":2336,"bytesExact":true,"sha256":"ee65bb901c876664ab8707c487157aa1a6ee57c65969b28fb5ec9dc211e68167","required":true},{"tensorFile":"configuration.json","description":"12Hz speech codec configuration","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-voice-design/speech_tokenizer","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-voice-design/speech_tokenizer/configuration.json","bytes":76,"bytesExact":true,"sha256":"6bc26d64eb5024b4d1dab5a52371958b429256d6c9d59787f1f5294a54e0cebd","required":true},{"tensorFile":"preprocessor_config.json","description":"12Hz speech codec preprocessor config","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-voice-design/speech_tokenizer","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-voice-design/speech_tokenizer/preprocessor_config.json","bytes":234,"bytesExact":true,"sha256":"fcb3805e597e786d4067706e602f6688524640f8d3396790e2e09b5942fcbdfb","required":true},{"tensorFile":"model.safetensors","description":"Qwen3-TTS 12Hz speech codec (encoder + decoder)","destinationFolder":"models/tts/qwen3-tts-12hz-1.7b-voice-design/speech_tokenizer","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/tts/qwen3-tts-12hz-1.7b-voice-design/speech_tokenizer/model.safetensors","bytes":682293092,"bytesExact":true,"sha256":"836b7b357f5ea43e889936a3709af68dfe3751881acefe4ecf0dbd30ba571258","required":true}]},"flashvsr_v1.1_tiny_long_bf16":{"description":"FlashVSR v1.1 Tiny Long BF16 standalone video upscaler (pinned JunhaoZhuang/FlashVSR-v1.1 revision 27561b18)","minVramGB":23,"files":[{"tensorFile":"diffusion_pytorch_model_streaming_dmd.safetensors","description":"FlashVSR v1.1 Tiny Long one-step streaming DMD transformer","destinationFolder":"models/FlashVSR-v1.1","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/FlashVSR-v1.1/diffusion_pytorch_model_streaming_dmd.safetensors","bytes":5676070392,"bytesExact":true,"sha256":"bd28180edcf3446c028e32fc6b731a80bf7e4da2ab4caac3186b9499964d37be","required":true},{"tensorFile":"LQ_proj_in.ckpt","description":"FlashVSR v1.1 low-quality video projection","destinationFolder":"models/FlashVSR-v1.1","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/FlashVSR-v1.1/LQ_proj_in.ckpt","bytes":575694948,"bytesExact":true,"sha256":"d6d011cdaaba6a52645086caa08fa04124e746f6ca568140a24007591142bfd2","required":true},{"tensorFile":"TCDecoder.ckpt","description":"FlashVSR v1.1 tiny conditional decoder","destinationFolder":"models/FlashVSR-v1.1","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/FlashVSR-v1.1/TCDecoder.ckpt","bytes":189018333,"bytesExact":true,"sha256":"e224bdcf2f52745cbf4d393ff5374c2ba09e90285d5d19062d2bf63b915b6161","required":true},{"tensorFile":"posi_prompt.pth","description":"FlashVSR v1.1 fixed positive prompt embedding","destinationFolder":"models/FlashVSR-v1.1","tensorDownload":"https://cdn.sogni.ai/ComfyUI/models/FlashVSR-v1.1/posi_prompt.pth","bytes":4195504,"bytesExact":true,"sha256":"4601107a11e4e11a936a6b79df579e54dbc99872132bf542151f0ffd65b4b1ef","required":true}]}},"workflowDependencies":{"minimax-h3-fl2va-fp8_t2v":["minimax-h3-shared","minimax-h3-fl2va-fp8"],"minimax-h3-fl2va-fp8_i2v":["minimax-h3-shared","minimax-h3-fl2va-fp8"],"minimax-h3-fl2va-fp8_flf2v":["minimax-h3-shared","minimax-h3-fl2va-fp8"],"minimax-h3-ref2va-fp8_r2v":["minimax-h3-shared","minimax-h3-ref2va-fp8"],"minimax-h3-fl2va-fp8_t2v_balanced":["minimax-h3-shared","minimax-h3-fl2va-fp8","minimax-h3-fl2va-balanced-lora"],"minimax-h3-fl2va-fp8_i2v_balanced":["minimax-h3-shared","minimax-h3-fl2va-fp8","minimax-h3-fl2va-balanced-lora"],"minimax-h3-fl2va-fp8_flf2v_balanced":["minimax-h3-shared","minimax-h3-fl2va-fp8","minimax-h3-fl2va-balanced-lora"],"minimax-h3-ref2va-fp8_r2v_balanced":["minimax-h3-shared","minimax-h3-ref2va-fp8","minimax-h3-ref2va-balanced-lora"],"minimax-h3-fl2va-fp8_t2v_turbo":["minimax-h3-shared","minimax-h3-fl2va-fp8","minimax-h3-fl2va-turbo-lora"],"minimax-h3-fl2va-fp8_i2v_turbo":["minimax-h3-shared","minimax-h3-fl2va-fp8","minimax-h3-fl2va-turbo-lora"],"minimax-h3-fl2va-fp8_flf2v_turbo":["minimax-h3-shared","minimax-h3-fl2va-fp8","minimax-h3-fl2va-turbo-lora"],"minimax-h3-fastvideo-int8_t2v_turbo":["minimax-h3-shared","minimax-h3-fastvideo-int8"],"minimax-h3-fastvideo-int8_i2v_turbo":["minimax-h3-shared","minimax-h3-fastvideo-int8"],"minimax-h3-fastvideo-int8_flf2v_turbo":["minimax-h3-shared","minimax-h3-fastvideo-int8"],"minimax-h3-fastvideo-int8_ia2v_turbo":["minimax-h3-shared","minimax-h3-fastvideo-int8"],"minimax-h3-fastvideo-int8_flfa2v_turbo":["minimax-h3-shared","minimax-h3-fastvideo-int8"],"minimax-h3-fastvideo-int8_a2v_turbo":["minimax-h3-shared","minimax-h3-fastvideo-int8"],"wan_v2.2-14b-fp8_t2v":["wan_v2.2-14b-fp8_shared","wan_v2.2-14b-fp8_t2v"],"wan_v2.2-14b-fp8_t2v_lightx2v":["wan_v2.2-14b-fp8_shared","wan_v2.2-14b-fp8_t2v","wan_v2.2-14b-fp8_t2v_lightx2v"],"wan_v2.2-14b-fp8_i2v":["wan_v2.2-14b-fp8_shared","wan_v2.2-14b-fp8_i2v"],"wan_v2.2-14b-fp8_i2v_lightx2v":["wan_v2.2-14b-fp8_shared","wan_v2.2-14b-fp8_i2v","wan_v2.2-14b-fp8_i2v_lightx2v"],"wan_v2.2-14b-fp8_s2v":["wan_v2.2-14b-fp8_shared","wan_v2.2-14b-fp8_s2v","wan_v2.2-14b-fp8_i2v"],"wan_v2.2-14b-fp8_s2v_lightx2v":["wan_v2.2-14b-fp8_shared","wan_v2.2-14b-fp8_s2v","wan_v2.2-14b-fp8_i2v","wan_v2.2-14b-fp8_i2v_lightx2v"],"wan_v2.2-14b-fp8_animate-move_lightx2v":["wan_v2.2-14b-fp8_shared","wan_v2.2-14b-fp8_animate","controlnet-aux-dwpose-torchscript","wan_v2.2-14b-fp8_animate_lightx2v","wan_v2.2-14b-fp8_i2v"],"wan_v2.2-14b-fp8_animate-replace_lightx2v":["wan_v2.2-14b-fp8_shared","wan_v2.2-14b-fp8_animate","controlnet-aux-dwpose-torchscript","wan_v2.2-14b-fp8_animate_lightx2v","wan_v2.2-14b-fp8_animate-replace_lightx2v","wan_v2.2-14b-fp8_i2v"],"z_image_turbo_bf16":["z_image_turbo_bf16"],"krea2_turbo_fp8_scaled":["krea2_turbo_fp8_scaled"],"krea2_identity_edit_v1_2":["krea2_turbo_fp8_scaled","krea2_identity_edit_v1_2"],"qwen_image_edit_2511_fp8":["qwen_image_edit_2511_fp8"],"qwen_image_edit_2511_fp8_lightning":["qwen_image_edit_2511_fp8","qwen_image_edit_2511_fp8_lightning"],"flux1-schnell-fp8":["flux1_shared","flux1-schnell-fp8"],"chroma-v.46-flash_fp8":["flux1_shared","chroma-v.46-flash_fp8"],"chroma-v48-detail-svd_fp8":["flux1_shared","chroma-v48-detail-svd_fp8"],"chroma1-hd_fp8_scaled":["flux1_shared","chroma1-hd_fp8_scaled"],"ltx23-22b-fp8_t2v_distilled":["ltx23-22b-fp8_shared","ltx23-22b-fp8","ltx23-22b-fp8_upscaler","ltx23-22b-fp8_id-lora-talkvid-3k"],"ltx23-22b-fp8_i2v_distilled":["ltx23-22b-fp8_shared","ltx23-22b-fp8","ltx23-22b-fp8_upscaler","ltx23-22b-fp8_id-lora-talkvid-3k"],"ltx23-22b-fp8_a2v_distilled":["ltx23-22b-fp8_shared","ltx23-22b-fp8","ltx23-22b-fp8_upscaler"],"ltx23-22b-fp8_ia2v_distilled":["ltx23-22b-fp8_shared","ltx23-22b-fp8","ltx23-22b-fp8_upscaler"],"ltx23-22b-fp8_t2v_dev":["ltx23-22b-fp8_shared","ltx23-22b-fp8_dev","ltx23-22b-fp8_distilled_lora","ltx23-22b-fp8_upscaler","ltx23-22b-fp8_id-lora-talkvid-3k"],"ltx23-22b-fp8_i2v_dev":["ltx23-22b-fp8_shared","ltx23-22b-fp8_dev","ltx23-22b-fp8_distilled_lora","ltx23-22b-fp8_upscaler","ltx23-22b-fp8_id-lora-talkvid-3k"],"ltx23-22b-fp8_a2v_dev":["ltx23-22b-fp8_shared","ltx23-22b-fp8_dev","ltx23-22b-fp8_distilled_lora","ltx23-22b-fp8_upscaler"],"ltx23-22b-fp8_ia2v_dev":["ltx23-22b-fp8_shared","ltx23-22b-fp8_dev","ltx23-22b-fp8_distilled_lora","ltx23-22b-fp8_upscaler"],"ltx23-22b-fp8_v2v_distilled":["ltx23-22b-fp8_shared","ltx23-22b-fp8","ltx23-22b-fp8_upscaler","ltx23-22b-fp8_ic-lora-union-control","ltx23-22b-fp8_ic-lora-in-outpaint","controlnet-aux-depth-anything-v2-small","controlnet-aux-yolox-torchscript","controlnet-aux-dwpose-torchscript"],"ltx23-22b-fp8_v2v_dev":["ltx23-22b-fp8_shared","ltx23-22b-fp8_dev","ltx23-22b-fp8_distilled_lora","ltx23-22b-fp8_upscaler","ltx23-22b-fp8_ic-lora-union-control","ltx23-22b-fp8_ic-lora-in-outpaint","controlnet-aux-depth-anything-v2-small","controlnet-aux-yolox-torchscript","controlnet-aux-dwpose-torchscript"],"ltx23-22b-10eros-v1.4-fp8mixed_i2v":["ltx23-22b-fp8mixed_10eros","ltx23-22b-fp8_text-projection","ltx23-22b-fp8_upscaler"],"ltx25-22b-int8_t2v_distilled":["ltx25-shared","ltx25-distilled","ltx25-upscaler"],"ltx25-22b-int8_i2v_distilled":["ltx25-shared","ltx25-distilled","ltx25-upscaler"],"ltx25-22b-int8_a2v_distilled":["ltx25-shared","ltx25-distilled","ltx25-upscaler"],"ltx25-22b-int8_ia2v_distilled":["ltx25-shared","ltx25-distilled","ltx25-upscaler"],"ltx25-22b-int8_v2v_distilled":["ltx25-shared","ltx25-distilled","ltx25-upscaler","ltx23-22b-fp8_ic-lora-union-control","ltx23-22b-fp8_ic-lora-in-outpaint","controlnet-aux-depth-anything-v2-small","controlnet-aux-yolox-torchscript","controlnet-aux-dwpose-torchscript"],"qwen_image_2512_fp8":["qwen_image_2512_fp8_shared","qwen_image_2512_fp8"],"qwen_image_2512_fp8_lightning":["qwen_image_2512_fp8_shared","qwen_image_2512_fp8","qwen_image_2512_lightning"],"z_image_bf16":["z_image_bf16_shared","z_image_bf16"],"ace_step_1.5_sft":["ace_step_1.5_shared","ace_step_1.5_sft"],"ace_step_1.5_turbo":["ace_step_1.5_shared","ace_step_1.5_turbo"],"ace_step_1.5_xl_sft":["ace_step_1.5_shared","ace_step_1.5_xl_sft"],"ace_step_1.5_xl_turbo":["ace_step_1.5_shared","ace_step_1.5_xl_turbo"],"dark_beast_z_image_turbo_v9_bf16":["z_image_bf16_shared","dark_beast_z_image_turbo_v9_bf16"],"dark_beast_krea2_fp8":["dark_beast_krea2_fp8"],"dark_beast_krea2_identity_edit_v1_2":["dark_beast_krea2_fp8","krea2_identity_edit_v1_2"],"one_obsession_v22_fp16":["one_obsession_v22_fp16"],"krea2_identity_edit_sogni_v0_3_alpha":["krea2_turbo_fp8_scaled","krea2_identity_edit_sogni_v0_3_alpha"],"minimax_music3":["minimax_music3"],"minimax-h3-ref2va-fp8_r2v_turbo":["minimax-h3-shared","minimax-h3-ref2va-fp8","minimax-h3-ref2va-turbo-lora"],"rtx_vsr_pro":[],"qwen3_tts_1.7b_custom_voice_bf16":["qwen3-tts-custom-voice"],"qwen3_tts_1.7b_voice_clone_bf16":["qwen3-tts-voice-clone"],"qwen3_tts_1.7b_voice_design_bf16":["qwen3-tts-voice-design"],"flashvsr_v1.1_tiny_long_bf16":["flashvsr_v1.1_tiny_long_bf16"]}}}