diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml index b17eafc..365e7d6 100644 --- a/.github/workflows/build.yml +++ b/.github/workflows/build.yml @@ -18,12 +18,12 @@ permissions: contents: read env: - LLAMA_CPP_VERSION: b10964 - LLAMA_CPP_MACOS_ARM64_SHA256: 033c845c1df9bf945ff37bb193238b40910b2244be3e1e637b2ceb5878f1a6f5 - LLAMA_CPP_WIN_VULKAN_X64_SHA256: 1ee3ad952f4ba71f438bd6d7bebef19e1c7af04adcaa35d08b4ddabb27d4c642 - LLAMA_CPP_WIN_CPU_ARM64_SHA256: 4b6a004b076eea47c318bea35cf1db2ff2bf037738b04645646ae8d7c3159478 - LLAMA_CPP_UBUNTU_VULKAN_X64_SHA256: 55d1e58e14c11eedea090bf088fdeefbfe7b4b09ee03bf6dba9834651769afcf - LLAMA_CPP_UBUNTU_VULKAN_ARM64_SHA256: f7864baa0edf5a059fb42c5efb5aceb96075aa1f41e6c3142b71ca69286cb0bb + LLAMA_CPP_VERSION: b11476 + LLAMA_CPP_MACOS_ARM64_SHA256: 577634a1b8a59e8dabe02ba10de1e610be0574dfaf1cf3020e6dd42853ed877e + LLAMA_CPP_WIN_VULKAN_X64_SHA256: 5c71e7b749697da4a8d46e9ee55486845cbba27c9dfbecb4007f31ba6610d523 + LLAMA_CPP_WIN_CPU_ARM64_SHA256: 68e3a218ed7d9cd563e8ddf7a1e58d88d034a8f90a91061bdd3876bf247d8a93 + LLAMA_CPP_UBUNTU_VULKAN_X64_SHA256: 5bb4306d7917f33e81efda02e6f791ae6a82e86bee121227a3ab2b4e8e40427f + LLAMA_CPP_UBUNTU_VULKAN_ARM64_SHA256: 8dae2f39afee01d3032a101d7c398a6734d6fd9697690d8f93bdce4e3a9efea2 jobs: test: diff --git a/README.md b/README.md index 1226e4b..1b23750 100644 --- a/README.md +++ b/README.md @@ -143,6 +143,13 @@ terms or store access tokens on a user's behalf. `--accept-license` records the local Tapioca acknowledgement; it does not approve access in the user's Hugging Face account. +Qwen Image 2.1 is available as `qwen-image-2.1` for Windows/Linux NVIDIA with +Diffusers 0.41. It supports image editing, native RGBA transparency, improved +text rendering, and multiple references. The roughly 31 GiB BF16 snapshot is +public but restricted to non-commercial research/evaluation by the Qwen +Research License, so Tapioca requires `--accept-license` without requiring an +Hugging Face token. + ## Everyday commands ```text diff --git a/catalog/catalog.json b/catalog/catalog.json index 6e699c9..8330ab2 100644 --- a/catalog/catalog.json +++ b/catalog/catalog.json @@ -1,7 +1,49 @@ { "schema": 1, - "generated_at": "2026-10-07T15:01:27.467Z", + "generated_at": "2026-10-08T02:35:16.907406Z", "models": { + "audio8-tts": { + "name": "audio8-tts", + "files": { + "0.6b": "", + "0.6b-mlx": "" + }, + "kind": "speech", + "backends": { + "0.6b": "speech-audio8", + "0.6b-mlx": "speech-audio8-mlx" + }, + "repos": { + "0.6b": "Audio8/Audio8-TTS-Preview-0.6b", + "0.6b-mlx": "mlx-community/Audio8-TTS-Preview-0.6b-bf16" + }, + "default": "0.6b", + "platform_defaults": { + "darwin/arm64": "0.6b-mlx" + }, + "sizes": { + "0.6b": "~2.4 GiB", + "0.6b-mlx": "~2.4 GiB" + }, + "memory": { + "0.6b": "8 GiB min; 12 GiB recommended", + "0.6b-mlx": "8 GiB min; 12 GiB recommended" + }, + "gpus": { + "0.6b": "NVIDIA CUDA recommended; CPU supported", + "0.6b-mlx": "Apple Silicon GPU" + }, + "languages": { + "0.6b": "11 languages", + "0.6b-mlx": "11 languages" + }, + "features": { + "0.6b": "preview, 44.1 kHz speech, zero-shot voice cloning, default voice", + "0.6b-mlx": "preview, 44.1 kHz speech, zero-shot voice cloning, Apple Silicon optimized" + }, + "license": "Apache-2.0", + "license_url": "https://huggingface.co/Audio8/Audio8-TTS-Preview-0.6b" + }, "chatterbox": { "name": "chatterbox", "repo": "ResembleAI/chatterbox", @@ -44,44 +86,44 @@ "name": "flux2-klein", "repo": "mlx-community/flux2-klein-4b-4bit", "files": { - "4b-q4-mlx": "", "4b-bf16-cuda": "", - "4b-fp8-cuda": "" + "4b-fp8-cuda": "", + "4b-q4-mlx": "" }, "kind": "image", "backends": { - "4b-q4-mlx": "mflux", "4b-bf16-cuda": "diffusers", - "4b-fp8-cuda": "diffusers" + "4b-fp8-cuda": "diffusers", + "4b-q4-mlx": "mflux" + }, + "repos": { + "4b-bf16-cuda": "black-forest-labs/FLUX.2-klein-4B", + "4b-fp8-cuda": "black-forest-labs/FLUX.2-klein-4b-fp8", + "4b-q4-mlx": "mlx-community/flux2-klein-4b-4bit" }, "default": "4b-q4-mlx", + "platform_defaults": { + "darwin/arm64": "4b-q4-mlx", + "linux": "4b-bf16-cuda", + "windows": "4b-bf16-cuda" + }, "width": 1024, "height": 1024, "steps": 4, "sizes": { - "4b-q4-mlx": "~4.6 GiB", "4b-bf16-cuda": "~15 GiB", - "4b-fp8-cuda": "~11 GiB" + "4b-fp8-cuda": "~11 GiB", + "4b-q4-mlx": "~4.6 GiB" }, "memory": { - "4b-q4-mlx": "16 GiB min; 24 GiB recommended", "4b-bf16-cuda": "24 GiB system RAM; 13 GiB VRAM min", - "4b-fp8-cuda": "24 GiB system RAM; 12 GiB VRAM recommended" + "4b-fp8-cuda": "24 GiB system RAM; 12 GiB VRAM recommended", + "4b-q4-mlx": "16 GiB min; 24 GiB recommended" }, "gpus": { - "4b-q4-mlx": "Apple Silicon GPU", "4b-bf16-cuda": "NVIDIA RTX 3090/4070 or newer", - "4b-fp8-cuda": "NVIDIA Ada or newer recommended; Ampere supported with reduced benefit" - }, - "repos": { - "4b-bf16-cuda": "black-forest-labs/FLUX.2-klein-4B", - "4b-fp8-cuda": "black-forest-labs/FLUX.2-klein-4b-fp8", - "4b-q4-mlx": "mlx-community/flux2-klein-4b-4bit" - }, - "platform_defaults": { - "darwin/arm64": "4b-q4-mlx", - "linux": "4b-bf16-cuda", - "windows": "4b-bf16-cuda" + "4b-fp8-cuda": "NVIDIA Ada or newer recommended; Ampere supported with reduced benefit", + "4b-q4-mlx": "Apple Silicon GPU" }, "features": { "4b-bf16-cuda": "4-step generation, image editing, multi-reference editing, Apache-2.0", @@ -212,11 +254,49 @@ "4b-q4_k_m": "~2.4 GiB" } }, + "gemma4": { + "name": "gemma4", + "repo": "ggml-org/gemma-4-E2B-it-GGUF", + "files": { + "e2b-mlx-4bit": "", + "e2b-q4_0": "gemma-4-E2B-it-Q4_0.gguf" + }, + "kind": "text", + "backends": { + "e2b-mlx-4bit": "mlx-vlm" + }, + "repos": { + "e2b-mlx-4bit": "mlx-community/gemma-4-e2b-it-4bit", + "e2b-q4_0": "ggml-org/gemma-4-E2B-it-GGUF" + }, + "default": "e2b-q4_0", + "platform_defaults": { + "darwin/arm64": "e2b-mlx-4bit" + }, + "sizes": { + "e2b-mlx-4bit": "~3.6 GiB", + "e2b-q4_0": "~2.8 GiB" + }, + "memory": { + "e2b-mlx-4bit": "12 GiB min; 16 GiB recommended", + "e2b-q4_0": "8 GiB min; 12 GiB recommended" + }, + "gpus": { + "e2b-mlx-4bit": "Apple Silicon GPU", + "e2b-q4_0": "Optional Metal, CUDA, Vulkan, or ROCm acceleration" + }, + "features": { + "e2b-mlx-4bit": "reasoning, system prompts, function calling, image understanding", + "e2b-q4_0": "reasoning, system prompts, function calling, 128K context" + }, + "license": "Apache-2.0", + "license_url": "https://huggingface.co/google/gemma-4-E2B-it" + }, "glm-4.7-flash": { "name": "glm-4.7-flash", "repo": "ggml-org/GLM-4.7-Flash-GGUF", "files": { - "q4_k_m": "GLM-4.7-Flash-Q4_K_M.gguf", + "q4_k_m": "GLM-4.7-Flash-Q4_K.gguf", "q8_0": "GLM-4.7-Flash-Q8_0.gguf" }, "default": "q4_k_m", @@ -329,235 +409,489 @@ "license": "Krea 2 Community License", "license_url": "https://huggingface.co/krea/Krea-2-Turbo" }, - "ltx-video": { - "name": "ltx-video", - "repo": "Lightricks/LTX-Video-0.9.5", + "ltx-2.5": { + "name": "ltx-2.5", + "repo": "Lightricks/LTX-2.5-Diffusers", "files": { - "2b-fp16": "" + "22b-bf16-cuda": "" }, "kind": "video", "backends": { - "2b-fp16": "diffusers-video" + "22b-bf16-cuda": "diffusers-ltx2" }, - "default": "2b-fp16", + "default": "22b-bf16-cuda", "width": 768, "height": 512, - "steps": 8, - "frames": 49, - "fps": 24, - "sizes": { - "2b-fp16": "~24 GiB" - }, - "memory": { - "2b-fp16": "24 GiB min; 32 GiB recommended" - }, - "gpus": { - "2b-fp16": "NVIDIA CUDA, 8 GiB+ VRAM; CPU offload" - } - }, - "minicpm5-2b": { - "context": 8192, - "name": "minicpm5-2b", - "repo": "openbmb/MiniCPM5-2B-GGUF", - "files": { - "q4_k_m": "MiniCPM5-2B-Q4_K_M.gguf" - }, - "default": "q4_k_m", - "sizes": { - "q4_k_m": "~1.45 GiB" - }, - "memory": { - "q4_k_m": "8 GiB test target; 16 GiB recommended at 8K context" - }, - "gpus": { - "q4_k_m": "CPU or Apple Metal; Vulkan pending qualification" - }, - "languages": { - "q4_k_m": "English and Chinese" - }, - "features": { - "q4_k_m": "basic text chat; tools and long context not yet qualified" - }, - "downloads": { - "q4_k_m": { - "revision": "2079a22f3beaa4e306449978533478fe0522f4b3", - "sha256": "ec2d5801640099e97d8d7e8003ad4d81f336e757811f03a26173dddf386602fd", - "size_bytes": 1561318368 - } - }, - "license": "Apache-2.0", - "license_url": "https://huggingface.co/openbmb/MiniCPM5-2B" - }, - "minimax-h3": { - "name": "minimax-h3", - "repo": "Comfy-Org/MiniMax-H3", - "files": { - "fl2va-int8-cuda": "", - "fl2va-int8-mac": "" - }, - "kind": "video", - "backends": { - "fl2va-int8-cuda": "comfy-h3-cuda", - "fl2va-int8-mac": "comfy-h3-mps" - }, - "default": "fl2va-int8-mac", - "width": 864, - "height": 480, - "steps": 20, - "frames": 73, + "steps": 30, + "frames": 121, "fps": 24, "sizes": { - "fl2va-int8-cuda": "~41 GiB", - "fl2va-int8-mac": "~41 GiB" + "22b-bf16-cuda": "~68 GiB" }, "memory": { - "fl2va-int8-cuda": "32 GiB system RAM; 16 GiB VRAM recommended", - "fl2va-int8-mac": "48 GiB min; 64 GiB recommended" + "22b-bf16-cuda": "96 GiB system RAM recommended; sequential CPU offload enabled" }, "gpus": { - "fl2va-int8-cuda": "NVIDIA CUDA, 16 GiB VRAM (4070 Ti SUPER supported)", - "fl2va-int8-mac": "Apple Silicon GPU" + "22b-bf16-cuda": "NVIDIA CUDA, 24 GiB VRAM minimum; 48 GiB recommended" }, "features": { - "fl2va-int8-cuda": "text/image-to-video, native stereo audio, LoRA stacks, managed ComfyUI", - "fl2va-int8-mac": "text/image-to-video, native stereo audio, LoRA stacks, managed ComfyUI" + "22b-bf16-cuda": "text-to-video, synchronized stereo audio, multishot prompting, 1-20 second clips" }, "artifacts": { - "fl2va-int8-cuda": [ + "22b-bf16-cuda": [ { - "repo": "Comfy-Org/MiniMax-H3", - "filename": "diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors", - "target": "diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors" + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "model_index.json", + "target": "model_index.json" }, { - "repo": "Comfy-Org/MiniMax-H3", - "filename": "text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors", - "target": "text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors" + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "modular_model_index.json", + "target": "modular_model_index.json" }, { - "repo": "Comfy-Org/MiniMax-H3", - "filename": "vae/minimax_h3_video_vae_fp16.safetensors", - "target": "vae/minimax_h3_video_vae_fp16.safetensors" + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "audio_vae/config.json", + "target": "audio_vae/config.json" }, { - "repo": "Comfy-Org/MiniMax-H3", - "filename": "vae/minimax_h3_audio_vae_fp32.safetensors", - "target": "vae/minimax_h3_audio_vae_fp32.safetensors" - } - ], - "fl2va-int8-mac": [ + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "audio_vae/diffusion_pytorch_model.safetensors", + "target": "audio_vae/diffusion_pytorch_model.safetensors" + }, { - "repo": "Comfy-Org/MiniMax-H3", - "filename": "diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors", - "target": "diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors" + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "connectors/config.json", + "target": "connectors/config.json" }, { - "repo": "realrebelai/MiniMax-H3_GGUFs", - "filename": "qwen3vl-32B-MiniMax-H3-Q4_K_M.gguf", - "target": "text_encoders/qwen3vl-32B-MiniMax-H3-Q4_K_M.gguf" + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "connectors/diffusion_pytorch_model.safetensors", + "target": "connectors/diffusion_pytorch_model.safetensors" }, { - "repo": "Comfy-Org/MiniMax-H3", - "filename": "vae/minimax_h3_video_vae_fp16.safetensors", - "target": "vae/minimax_h3_video_vae_fp16.safetensors" + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "diffusion_decoder/config.json", + "target": "diffusion_decoder/config.json" }, { - "repo": "Comfy-Org/MiniMax-H3", - "filename": "vae/minimax_h3_audio_vae_fp32.safetensors", - "target": "vae/minimax_h3_audio_vae_fp32.safetensors" - } - ] - } - }, - "phi4-mini": { - "name": "phi4-mini", - "repo": "unsloth/Phi-4-mini-instruct-GGUF", - "files": { - "q4_k_m": "Phi-4-mini-instruct-Q4_K_M.gguf", - "q5_k_m": "Phi-4-mini-instruct-Q5_K_M.gguf" - }, - "kind": "text", - "default": "q4_k_m", - "sizes": { - "q4_k_m": "~2.4 GiB", - "q5_k_m": "~2.7 GiB" - } + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "diffusion_decoder/diffusion_pytorch_model.safetensors", + "target": "diffusion_decoder/diffusion_pytorch_model.safetensors" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "duration_head/config.json", + "target": "duration_head/config.json" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "duration_head/diffusion_pytorch_model.safetensors", + "target": "duration_head/diffusion_pytorch_model.safetensors" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "latent_upsampler/config.json", + "target": "latent_upsampler/config.json" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "latent_upsampler/diffusion_pytorch_model.safetensors", + "target": "latent_upsampler/diffusion_pytorch_model.safetensors" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "scheduler/scheduler_config.json", + "target": "scheduler/scheduler_config.json" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "temporal_latent_upsampler/config.json", + "target": "temporal_latent_upsampler/config.json" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "temporal_latent_upsampler/diffusion_pytorch_model.safetensors", + "target": "temporal_latent_upsampler/diffusion_pytorch_model.safetensors" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "text_encoder/config.json", + "target": "text_encoder/config.json" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "text_encoder/generation_config.json", + "target": "text_encoder/generation_config.json" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "text_encoder/model-00001-of-00005.safetensors", + "target": "text_encoder/model-00001-of-00005.safetensors" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "text_encoder/model-00002-of-00005.safetensors", + "target": "text_encoder/model-00002-of-00005.safetensors" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "text_encoder/model-00003-of-00005.safetensors", + "target": "text_encoder/model-00003-of-00005.safetensors" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "text_encoder/model-00004-of-00005.safetensors", + "target": "text_encoder/model-00004-of-00005.safetensors" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "text_encoder/model-00005-of-00005.safetensors", + "target": "text_encoder/model-00005-of-00005.safetensors" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "text_encoder/model.safetensors.index.json", + "target": "text_encoder/model.safetensors.index.json" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "tokenizer/chat_template.jinja", + "target": "tokenizer/chat_template.jinja" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "tokenizer/tokenizer.json", + "target": "tokenizer/tokenizer.json" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "tokenizer/tokenizer_config.json", + "target": "tokenizer/tokenizer_config.json" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "vae/config.json", + "target": "vae/config.json" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "vae/diffusion_pytorch_model.safetensors", + "target": "vae/diffusion_pytorch_model.safetensors" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "vocoder/config.json", + "target": "vocoder/config.json" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "vocoder/diffusion_pytorch_model.safetensors", + "target": "vocoder/diffusion_pytorch_model.safetensors" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "transformer_full/config.json", + "target": "transformer/config.json" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "transformer_full/diffusion_pytorch_model-00001-of-00004.safetensors", + "target": "transformer/diffusion_pytorch_model-00001-of-00004.safetensors" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "transformer_full/diffusion_pytorch_model-00002-of-00004.safetensors", + "target": "transformer/diffusion_pytorch_model-00002-of-00004.safetensors" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "transformer_full/diffusion_pytorch_model-00003-of-00004.safetensors", + "target": "transformer/diffusion_pytorch_model-00003-of-00004.safetensors" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "transformer_full/diffusion_pytorch_model-00004-of-00004.safetensors", + "target": "transformer/diffusion_pytorch_model-00004-of-00004.safetensors" + }, + { + "repo": "Lightricks/LTX-2.5-Diffusers", + "filename": "transformer_full/diffusion_pytorch_model.safetensors.index.json", + "target": "transformer/diffusion_pytorch_model.safetensors.index.json" + } + ] + }, + "gated": true, + "license": "LTX-2 Community License", + "license_url": "https://huggingface.co/Lightricks/LTX-2.5-Diffusers" }, - "qwen-image-flash": { - "name": "qwen-image-flash", - "repo": "mlx-community/Qwen-Image-Flash-8bit", + "ltx-video": { + "name": "ltx-video", + "repo": "Lightricks/LTX-Video-0.9.5", "files": { - "8bit": "", - "bf16": "", - "int8": "" + "2b-fp16": "" }, - "kind": "image", + "kind": "video", "backends": { - "8bit": "mlx", - "bf16": "diffusers", - "int8": "mlx" + "2b-fp16": "diffusers-video" }, - "default": "int8", - "width": 1024, - "height": 1024, - "steps": 4, + "default": "2b-fp16", + "width": 768, + "height": 512, + "steps": 8, + "frames": 49, + "fps": 24, "sizes": { - "8bit": "~28 GiB", - "bf16": "~58 GiB", - "int8": "~28 GiB" + "2b-fp16": "~24 GiB" }, "memory": { - "8bit": "48 GiB min; 64 GiB recommended", - "bf16": "64 GiB min; 96 GiB recommended", - "int8": "48 GiB min; 64 GiB recommended" + "2b-fp16": "24 GiB min; 32 GiB recommended" }, "gpus": { - "8bit": "Apple Silicon GPU", - "bf16": "NVIDIA Ampere+, 16 GiB VRAM with offload", - "int8": "Apple Silicon GPU" + "2b-fp16": "NVIDIA CUDA, 8 GiB+ VRAM; CPU offload" } }, - "qwen3": { - "name": "qwen3", - "repo": "Qwen/Qwen3-8B-GGUF", + "minicpm5-2b": { + "context": 8192, + "name": "minicpm5-2b", + "repo": "openbmb/MiniCPM5-2B-GGUF", "files": { - "30b-mlx": "", - "4b-q4_k_m": "Qwen3-4B-Q4_K_M.gguf", - "4b-q5_k_m": "Qwen3-4B-Q5_K_M.gguf", - "8b-q4_k_m": "Qwen3-8B-Q4_K_M.gguf", - "8b-q5_k_m": "Qwen3-8B-Q5_K_M.gguf" + "q4_k_m": "MiniCPM5-2B-Q4_K_M.gguf" }, - "kind": "text", - "backends": { - "30b-mlx": "mlx-vlm" + "default": "q4_k_m", + "sizes": { + "q4_k_m": "~1.45 GiB" }, - "repos": { - "30b-mlx": "mlx-community/Qwen3-30B-A3B-4bit", - "4b-q4_k_m": "Qwen/Qwen3-4B-GGUF", - "4b-q5_k_m": "Qwen/Qwen3-4B-GGUF", - "8b-q4_k_m": "Qwen/Qwen3-8B-GGUF", - "8b-q5_k_m": "Qwen/Qwen3-8B-GGUF" + "memory": { + "q4_k_m": "8 GiB test target; 16 GiB recommended at 8K context" }, - "default": "8b-q4_k_m", - "sizes": { - "30b-mlx": "~16 GiB", - "4b-q4_k_m": "~2.4 GiB", - "4b-q5_k_m": "~2.8 GiB", - "8b-q4_k_m": "~4.7 GiB", - "8b-q5_k_m": "~5.5 GiB" - } + "gpus": { + "q4_k_m": "CPU or Apple Metal; Vulkan pending qualification" + }, + "languages": { + "q4_k_m": "English and Chinese" + }, + "features": { + "q4_k_m": "basic text chat; tools and long context not yet qualified" + }, + "downloads": { + "q4_k_m": { + "revision": "2079a22f3beaa4e306449978533478fe0522f4b3", + "sha256": "ec2d5801640099e97d8d7e8003ad4d81f336e757811f03a26173dddf386602fd", + "size_bytes": 1561318368 + } + }, + "license": "Apache-2.0", + "license_url": "https://huggingface.co/openbmb/MiniCPM5-2B" }, - "qwen3-coder": { - "name": "qwen3-coder", + "minimax-h3": { + "name": "minimax-h3", + "repo": "Comfy-Org/MiniMax-H3", "files": { - "30b-mlx": "" + "fl2va-int8-cuda": "", + "fl2va-int8-mac": "" }, - "kind": "text", + "kind": "video", "backends": { - "30b-mlx": "mlx-vlm" + "fl2va-int8-cuda": "comfy-h3-cuda", + "fl2va-int8-mac": "comfy-h3-mps" }, - "repos": { - "30b-mlx": "mlx-community/Qwen3-Coder-30B-A3B-Instruct-4bit" + "default": "fl2va-int8-mac", + "width": 864, + "height": 480, + "steps": 20, + "frames": 73, + "fps": 24, + "sizes": { + "fl2va-int8-cuda": "~41 GiB", + "fl2va-int8-mac": "~41 GiB" + }, + "memory": { + "fl2va-int8-cuda": "32 GiB system RAM; 16 GiB VRAM recommended", + "fl2va-int8-mac": "48 GiB min; 64 GiB recommended" + }, + "gpus": { + "fl2va-int8-cuda": "NVIDIA CUDA, 16 GiB VRAM (4070 Ti SUPER supported)", + "fl2va-int8-mac": "Apple Silicon GPU" + }, + "features": { + "fl2va-int8-cuda": "text/image-to-video, native stereo audio, LoRA stacks, managed ComfyUI", + "fl2va-int8-mac": "text/image-to-video, native stereo audio, LoRA stacks, managed ComfyUI" + }, + "artifacts": { + "fl2va-int8-cuda": [ + { + "repo": "Comfy-Org/MiniMax-H3", + "filename": "diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors", + "target": "diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors" + }, + { + "repo": "Comfy-Org/MiniMax-H3", + "filename": "text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors", + "target": "text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors" + }, + { + "repo": "Comfy-Org/MiniMax-H3", + "filename": "vae/minimax_h3_video_vae_fp16.safetensors", + "target": "vae/minimax_h3_video_vae_fp16.safetensors" + }, + { + "repo": "Comfy-Org/MiniMax-H3", + "filename": "vae/minimax_h3_audio_vae_fp32.safetensors", + "target": "vae/minimax_h3_audio_vae_fp32.safetensors" + } + ], + "fl2va-int8-mac": [ + { + "repo": "Comfy-Org/MiniMax-H3", + "filename": "diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors", + "target": "diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors" + }, + { + "repo": "realrebelai/MiniMax-H3_GGUFs", + "filename": "qwen3vl-32B-MiniMax-H3-Q4_K_M.gguf", + "target": "text_encoders/qwen3vl-32B-MiniMax-H3-Q4_K_M.gguf" + }, + { + "repo": "Comfy-Org/MiniMax-H3", + "filename": "vae/minimax_h3_video_vae_fp16.safetensors", + "target": "vae/minimax_h3_video_vae_fp16.safetensors" + }, + { + "repo": "Comfy-Org/MiniMax-H3", + "filename": "vae/minimax_h3_audio_vae_fp32.safetensors", + "target": "vae/minimax_h3_audio_vae_fp32.safetensors" + } + ] + } + }, + "phi4-mini": { + "name": "phi4-mini", + "repo": "unsloth/Phi-4-mini-instruct-GGUF", + "files": { + "q4_k_m": "Phi-4-mini-instruct-Q4_K_M.gguf", + "q5_k_m": "Phi-4-mini-instruct-Q5_K_M.gguf" + }, + "kind": "text", + "default": "q4_k_m", + "sizes": { + "q4_k_m": "~2.4 GiB", + "q5_k_m": "~2.7 GiB" + } + }, + "qwen-image-2.1": { + "name": "qwen-image-2.1", + "repo": "Qwen/Qwen-Image-2.1", + "files": { + "bf16-cuda": "" + }, + "kind": "image", + "backends": { + "bf16-cuda": "diffusers" + }, + "default": "bf16-cuda", + "width": 1024, + "height": 1024, + "steps": 40, + "sizes": { + "bf16-cuda": "~31 GiB" + }, + "memory": { + "bf16-cuda": "48 GiB min; 64 GiB recommended" + }, + "gpus": { + "bf16-cuda": "NVIDIA CUDA, 24 GiB+ VRAM; CPU offload recommended" + }, + "languages": { + "bf16-cuda": "multilingual prompts and typography" + }, + "features": { + "bf16-cuda": "text-to-image, image editing, native RGBA transparency, up to 10 references" + }, + "downloads": { + "bf16-cuda": { + "revision": "d26bb61231c349cf6b7896fa83353113880e1ba3" + } + }, + "license_acceptance_required": true, + "license": "Qwen Research License Agreement (non-commercial only)", + "license_url": "https://huggingface.co/Qwen/Qwen-Image-2.1/blob/main/LICENSE" + }, + "qwen-image-flash": { + "name": "qwen-image-flash", + "repo": "mlx-community/Qwen-Image-Flash-8bit", + "files": { + "8bit": "", + "bf16": "", + "int8": "" + }, + "kind": "image", + "backends": { + "8bit": "mlx", + "bf16": "diffusers", + "int8": "mlx" + }, + "default": "int8", + "width": 1024, + "height": 1024, + "steps": 4, + "sizes": { + "8bit": "~28 GiB", + "bf16": "~58 GiB", + "int8": "~28 GiB" + }, + "memory": { + "8bit": "48 GiB min; 64 GiB recommended", + "bf16": "64 GiB min; 96 GiB recommended", + "int8": "48 GiB min; 64 GiB recommended" + }, + "gpus": { + "8bit": "Apple Silicon GPU", + "bf16": "NVIDIA Ampere+, 16 GiB VRAM with offload", + "int8": "Apple Silicon GPU" + } + }, + "qwen3": { + "name": "qwen3", + "repo": "Qwen/Qwen3-8B-GGUF", + "files": { + "30b-mlx": "", + "4b-q4_k_m": "Qwen3-4B-Q4_K_M.gguf", + "4b-q5_k_m": "Qwen3-4B-Q5_K_M.gguf", + "8b-q4_k_m": "Qwen3-8B-Q4_K_M.gguf", + "8b-q5_k_m": "Qwen3-8B-Q5_K_M.gguf" + }, + "kind": "text", + "backends": { + "30b-mlx": "mlx-vlm" + }, + "repos": { + "30b-mlx": "mlx-community/Qwen3-30B-A3B-4bit", + "4b-q4_k_m": "Qwen/Qwen3-4B-GGUF", + "4b-q5_k_m": "Qwen/Qwen3-4B-GGUF", + "8b-q4_k_m": "Qwen/Qwen3-8B-GGUF", + "8b-q5_k_m": "Qwen/Qwen3-8B-GGUF" + }, + "default": "8b-q4_k_m", + "sizes": { + "30b-mlx": "~16 GiB", + "4b-q4_k_m": "~2.4 GiB", + "4b-q5_k_m": "~2.8 GiB", + "8b-q4_k_m": "~4.7 GiB", + "8b-q5_k_m": "~5.5 GiB" + } + }, + "qwen3-coder": { + "name": "qwen3-coder", + "files": { + "30b-mlx": "" + }, + "kind": "text", + "backends": { + "30b-mlx": "mlx-vlm" + }, + "repos": { + "30b-mlx": "mlx-community/Qwen3-Coder-30B-A3B-Instruct-4bit" }, "default": "30b-mlx", "sizes": { @@ -654,563 +988,189 @@ "qwen3.6": { "name": "qwen3.6", "files": { - "35b-mlx": "", - "35b-mlx-4bit": "", - "35b-mlx-6bit": "", - "35b-mlx-8bit": "", "27b-q4_k_m": "Qwen3.6-27B-Q4_K_M.gguf", - "27b-q8_0": "Qwen3.6-27B-Q8_0.gguf" - }, - "kind": "text", - "backends": { - "35b-mlx": "mlx-vlm", - "35b-mlx-4bit": "mlx-vlm", - "35b-mlx-6bit": "mlx-vlm", - "35b-mlx-8bit": "mlx-vlm" - }, - "repos": { - "35b-mlx": "mlx-community/Qwen3.6-35B-A3B-4bit", - "35b-mlx-4bit": "mlx-community/Qwen3.6-35B-A3B-4bit", - "35b-mlx-6bit": "mlx-community/Qwen3.6-35B-A3B-6bit", - "35b-mlx-8bit": "mlx-community/Qwen3.6-35B-A3B-8bit", - "27b-q4_k_m": "ggml-org/Qwen3.6-27B-GGUF", - "27b-q8_0": "ggml-org/Qwen3.6-27B-GGUF" - }, - "default": "35b-mlx", - "sizes": { - "35b-mlx": "~20 GiB", - "35b-mlx-4bit": "~20 GiB", - "35b-mlx-6bit": "~27 GiB", - "35b-mlx-8bit": "~36 GiB", - "27b-q4_k_m": "~19.1 GiB", - "27b-q8_0": "~28.6 GiB" - }, - "platform_defaults": { - "darwin/arm64": "35b-mlx", - "linux": "27b-q4_k_m", - "windows": "27b-q4_k_m" - }, - "memory": { - "27b-q4_k_m": "24 GiB min; 32 GiB recommended", - "27b-q8_0": "40 GiB min; 48 GiB recommended" - }, - "gpus": { - "27b-q4_k_m": "Optional Metal, CUDA, Vulkan, or ROCm acceleration", - "27b-q8_0": "Optional Metal, CUDA, Vulkan, or ROCm acceleration" - }, - "features": { - "27b-q4_k_m": "reasoning, coding, tool use, image understanding with optional mmproj", - "27b-q8_0": "higher-fidelity reasoning, coding, tool use, image understanding with optional mmproj" - } - }, - "qwen3.8": { - "name": "qwen3.8", - "repo": "unsloth/Qwen3.8-27B-GGUF", - "files": { - "27b-mlx": "", - "27b-q4_k_m": "Qwen3.8-27B-UD-Q4_K_M.gguf", - "27b-q5_k_m": "Qwen3.8-27B-UD-Q5_K_M.gguf" - }, - "kind": "text", - "backends": { - "27b-mlx": "mlx-vlm" - }, - "repos": { - "27b-mlx": "mlx-community/Qwen3.8-27B-4bit" - }, - "default": "27b-q4_k_m", - "platform_defaults": { - "darwin/arm64": "27b-mlx" - }, - "sizes": { - "27b-mlx": "~15 GiB", - "27b-q4_k_m": "~16 GiB", - "27b-q5_k_m": "~19 GiB" - }, - "memory": { - "27b-mlx": "24 GiB min; 32 GiB recommended", - "27b-q4_k_m": "24 GiB min; 32 GiB recommended", - "27b-q5_k_m": "32 GiB min; 40 GiB recommended" - }, - "gpus": { - "27b-mlx": "Apple Silicon GPU", - "27b-q4_k_m": "Optional GPU acceleration", - "27b-q5_k_m": "Optional GPU acceleration" - }, - "features": { - "27b-mlx": "agentic coding, tool use, Apple Silicon optimized", - "27b-q4_k_m": "agentic coding, tool use, long-horizon reasoning", - "27b-q5_k_m": "agentic coding, tool use, long-horizon reasoning" - } - }, - "sd-turbo": { - "name": "sd-turbo", - "repo": "stabilityai/sd-turbo", - "files": { - "fp16": "", - "onnx-arm64": "", - "onnx-directml": "" - }, - "kind": "image", - "backends": { - "fp16": "diffusers", - "onnx-arm64": "onnx-cpu", - "onnx-directml": "onnx-directml" - }, - "repos": { - "onnx-arm64": "Heliosoph/sd-turbo-onnx", - "onnx-directml": "Heliosoph/sd-turbo-onnx" - }, - "default": "fp16", - "width": 512, - "height": 512, - "steps": 4, - "sizes": { - "fp16": "~3 GiB", - "onnx-arm64": "~4.8 GiB", - "onnx-directml": "~4.8 GiB" - }, - "memory": { - "fp16": "16 GiB min; 24 GiB recommended", - "onnx-arm64": "12 GiB min; 16 GiB recommended", - "onnx-directml": "12 GiB min; 16 GiB recommended" - }, - "gpus": { - "fp16": "NVIDIA CUDA, 6 GiB+ VRAM", - "onnx-arm64": "CPU (native Windows ARM64)", - "onnx-directml": "AMD, Intel, or NVIDIA DirectX 12 GPU" - } - }, - "sdxl-turbo": { - "name": "sdxl-turbo", - "repo": "stabilityai/sdxl-turbo", - "files": { - "fp16": "" - }, - "kind": "image", - "backends": { - "fp16": "diffusers" - }, - "default": "fp16", - "width": 1024, - "height": 1024, - "steps": 4, - "sizes": { - "fp16": "~7 GiB" - }, - "memory": { - "fp16": "24 GiB min; 32 GiB recommended" - }, - "gpus": { - "fp16": "NVIDIA CUDA, 8 GiB+ VRAM" - } - }, - "spark-x2.5-4b": { - "context": 8192, - "name": "spark-x2.5-4b", - "repo": "XHToken/Spark-X2.5-4B-GGUF", - "files": { - "q4_k_m": "Spark-X2.5-4B-Q4_K_M.gguf" - }, - "default": "q4_k_m", - "sizes": { - "q4_k_m": "~2.42 GiB" - }, - "memory": { - "q4_k_m": "8 GiB test target; 16 GiB recommended at 8K context" - }, - "gpus": { - "q4_k_m": "CPU or Apple Metal; Vulkan pending qualification" - }, - "languages": { - "q4_k_m": "multilingual" - }, - "features": { - "q4_k_m": "basic text chat; tools and long context not yet qualified" - }, - "downloads": { - "q4_k_m": { - "revision": "902d865994943ab9235670e24f01846ee06091f2", - "sha256": "adfcfa19a4ed6a5985da8bf565fe15f8e1a7e131d79bae2d19d48d1c40109428", - "size_bytes": 2600224352 - } - }, - "license": "Apache-2.0", - "license_url": "https://huggingface.co/XHToken/Spark-X2.5-4B" - }, - "stable-video-diffusion": { - "name": "stable-video-diffusion", - "repo": "stabilityai/stable-video-diffusion-img2vid-xt", - "files": { - "xt-fp16": "" - }, - "kind": "video", - "backends": { - "xt-fp16": "diffusers-video" - }, - "default": "xt-fp16", - "width": 1024, - "height": 576, - "steps": 25, - "frames": 25, - "fps": 7, - "sizes": { - "xt-fp16": "~4.5 GiB" - }, - "memory": { - "xt-fp16": "16 GiB min; 24 GiB recommended" - }, - "gpus": { - "xt-fp16": "NVIDIA CUDA, 8 GiB+ VRAM; CPU offload" - } - }, - "wan2.2-video": { - "name": "wan2.2-video", - "repo": "Anes1032/Wan2.2-TI2V-5B-mlx-q8", - "files": { - "5b-q8-mlx": "" - }, - "kind": "video", - "backends": { - "5b-q8-mlx": "mlx-video" - }, - "default": "5b-q8-mlx", - "width": 832, - "height": 480, - "steps": 40, - "frames": 41, - "fps": 24, - "sizes": { - "5b-q8-mlx": "~18 GiB" - }, - "memory": { - "5b-q8-mlx": "32 GiB min; 48 GiB recommended" - }, - "gpus": { - "5b-q8-mlx": "Apple Silicon GPU" - } - }, - "yume-video": { - "name": "yume-video", - "repo": "ckurasek/Yume-1.5-5B-720P-MLX", - "files": { - "5b-mlx": "" - }, - "kind": "video", - "backends": { - "5b-mlx": "mlx-video" - }, - "default": "5b-mlx", - "width": 832, - "height": 480, - "steps": 30, - "frames": 81, - "fps": 24, - "sizes": { - "5b-mlx": "~23 GiB" - }, - "memory": { - "5b-mlx": "32 GiB min; 48 GiB recommended" - }, - "gpus": { - "5b-mlx": "Apple Silicon GPU" - } - }, - "audio8-tts": { - "name": "audio8-tts", - "files": { - "0.6b": "", - "0.6b-mlx": "" - }, - "kind": "speech", - "backends": { - "0.6b": "speech-audio8", - "0.6b-mlx": "speech-audio8-mlx" - }, - "repos": { - "0.6b": "Audio8/Audio8-TTS-Preview-0.6b", - "0.6b-mlx": "mlx-community/Audio8-TTS-Preview-0.6b-bf16" - }, - "default": "0.6b", - "platform_defaults": { - "darwin/arm64": "0.6b-mlx" - }, - "sizes": { - "0.6b": "~2.4 GiB", - "0.6b-mlx": "~2.4 GiB" - }, - "memory": { - "0.6b": "8 GiB min; 12 GiB recommended", - "0.6b-mlx": "8 GiB min; 12 GiB recommended" - }, - "gpus": { - "0.6b": "NVIDIA CUDA recommended; CPU supported", - "0.6b-mlx": "Apple Silicon GPU" - }, - "languages": { - "0.6b": "11 languages", - "0.6b-mlx": "11 languages" - }, - "features": { - "0.6b": "preview, 44.1 kHz speech, zero-shot voice cloning, default voice", - "0.6b-mlx": "preview, 44.1 kHz speech, zero-shot voice cloning, Apple Silicon optimized" - }, - "license": "Apache-2.0", - "license_url": "https://huggingface.co/Audio8/Audio8-TTS-Preview-0.6b" - }, - "gemma4": { - "name": "gemma4", - "repo": "ggml-org/gemma-4-E2B-it-GGUF", - "files": { - "e2b-mlx-4bit": "", - "e2b-q4_0": "gemma-4-E2B-it-Q4_0.gguf" - }, - "kind": "text", - "backends": { - "e2b-mlx-4bit": "mlx-vlm" - }, - "repos": { - "e2b-mlx-4bit": "mlx-community/gemma-4-e2b-it-4bit", - "e2b-q4_0": "ggml-org/gemma-4-E2B-it-GGUF" - }, - "default": "e2b-q4_0", - "platform_defaults": { - "darwin/arm64": "e2b-mlx-4bit" - }, - "sizes": { - "e2b-mlx-4bit": "~3.6 GiB", - "e2b-q4_0": "~2.8 GiB" - }, - "memory": { - "e2b-mlx-4bit": "12 GiB min; 16 GiB recommended", - "e2b-q4_0": "8 GiB min; 12 GiB recommended" - }, - "gpus": { - "e2b-mlx-4bit": "Apple Silicon GPU", - "e2b-q4_0": "Optional Metal, CUDA, Vulkan, or ROCm acceleration" - }, - "features": { - "e2b-mlx-4bit": "reasoning, system prompts, function calling, image understanding", - "e2b-q4_0": "reasoning, system prompts, function calling, 128K context" - }, - "license": "Apache-2.0", - "license_url": "https://huggingface.co/google/gemma-4-E2B-it" - }, - "ltx-2.5": { - "name": "ltx-2.5", - "repo": "Lightricks/LTX-2.5-Diffusers", - "files": { - "22b-bf16-cuda": "" - }, - "kind": "video", - "backends": { - "22b-bf16-cuda": "diffusers-ltx2" - }, - "default": "22b-bf16-cuda", - "width": 768, - "height": 512, - "steps": 30, - "frames": 121, - "fps": 24, - "sizes": { - "22b-bf16-cuda": "~68 GiB" - }, - "memory": { - "22b-bf16-cuda": "96 GiB system RAM recommended; sequential CPU offload enabled" - }, - "gpus": { - "22b-bf16-cuda": "NVIDIA CUDA, 24 GiB VRAM minimum; 48 GiB recommended" - }, - "features": { - "22b-bf16-cuda": "text-to-video, synchronized stereo audio, multishot prompting, 1-20 second clips" - }, - "artifacts": { - "22b-bf16-cuda": [ - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "model_index.json", - "target": "model_index.json" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "modular_model_index.json", - "target": "modular_model_index.json" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "audio_vae/config.json", - "target": "audio_vae/config.json" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "audio_vae/diffusion_pytorch_model.safetensors", - "target": "audio_vae/diffusion_pytorch_model.safetensors" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "connectors/config.json", - "target": "connectors/config.json" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "connectors/diffusion_pytorch_model.safetensors", - "target": "connectors/diffusion_pytorch_model.safetensors" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "diffusion_decoder/config.json", - "target": "diffusion_decoder/config.json" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "diffusion_decoder/diffusion_pytorch_model.safetensors", - "target": "diffusion_decoder/diffusion_pytorch_model.safetensors" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "duration_head/config.json", - "target": "duration_head/config.json" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "duration_head/diffusion_pytorch_model.safetensors", - "target": "duration_head/diffusion_pytorch_model.safetensors" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "latent_upsampler/config.json", - "target": "latent_upsampler/config.json" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "latent_upsampler/diffusion_pytorch_model.safetensors", - "target": "latent_upsampler/diffusion_pytorch_model.safetensors" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "scheduler/scheduler_config.json", - "target": "scheduler/scheduler_config.json" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "temporal_latent_upsampler/config.json", - "target": "temporal_latent_upsampler/config.json" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "temporal_latent_upsampler/diffusion_pytorch_model.safetensors", - "target": "temporal_latent_upsampler/diffusion_pytorch_model.safetensors" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "text_encoder/config.json", - "target": "text_encoder/config.json" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "text_encoder/generation_config.json", - "target": "text_encoder/generation_config.json" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "text_encoder/model-00001-of-00005.safetensors", - "target": "text_encoder/model-00001-of-00005.safetensors" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "text_encoder/model-00002-of-00005.safetensors", - "target": "text_encoder/model-00002-of-00005.safetensors" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "text_encoder/model-00003-of-00005.safetensors", - "target": "text_encoder/model-00003-of-00005.safetensors" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "text_encoder/model-00004-of-00005.safetensors", - "target": "text_encoder/model-00004-of-00005.safetensors" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "text_encoder/model-00005-of-00005.safetensors", - "target": "text_encoder/model-00005-of-00005.safetensors" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "text_encoder/model.safetensors.index.json", - "target": "text_encoder/model.safetensors.index.json" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "tokenizer/chat_template.jinja", - "target": "tokenizer/chat_template.jinja" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "tokenizer/tokenizer.json", - "target": "tokenizer/tokenizer.json" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "tokenizer/tokenizer_config.json", - "target": "tokenizer/tokenizer_config.json" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "vae/config.json", - "target": "vae/config.json" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "vae/diffusion_pytorch_model.safetensors", - "target": "vae/diffusion_pytorch_model.safetensors" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "vocoder/config.json", - "target": "vocoder/config.json" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "vocoder/diffusion_pytorch_model.safetensors", - "target": "vocoder/diffusion_pytorch_model.safetensors" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "transformer_full/config.json", - "target": "transformer/config.json" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "transformer_full/diffusion_pytorch_model-00001-of-00004.safetensors", - "target": "transformer/diffusion_pytorch_model-00001-of-00004.safetensors" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "transformer_full/diffusion_pytorch_model-00002-of-00004.safetensors", - "target": "transformer/diffusion_pytorch_model-00002-of-00004.safetensors" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "transformer_full/diffusion_pytorch_model-00003-of-00004.safetensors", - "target": "transformer/diffusion_pytorch_model-00003-of-00004.safetensors" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "transformer_full/diffusion_pytorch_model-00004-of-00004.safetensors", - "target": "transformer/diffusion_pytorch_model-00004-of-00004.safetensors" - }, - { - "repo": "Lightricks/LTX-2.5-Diffusers", - "filename": "transformer_full/diffusion_pytorch_model.safetensors.index.json", - "target": "transformer/diffusion_pytorch_model.safetensors.index.json" - } - ] + "27b-q8_0": "Qwen3.6-27B-Q8_0.gguf", + "35b-mlx": "", + "35b-mlx-4bit": "", + "35b-mlx-6bit": "", + "35b-mlx-8bit": "" }, - "gated": true, - "license": "LTX-2 Community License", - "license_url": "https://huggingface.co/Lightricks/LTX-2.5-Diffusers" + "kind": "text", + "backends": { + "35b-mlx": "mlx-vlm", + "35b-mlx-4bit": "mlx-vlm", + "35b-mlx-6bit": "mlx-vlm", + "35b-mlx-8bit": "mlx-vlm" + }, + "repos": { + "27b-q4_k_m": "ggml-org/Qwen3.6-27B-GGUF", + "27b-q8_0": "ggml-org/Qwen3.6-27B-GGUF", + "35b-mlx": "mlx-community/Qwen3.6-35B-A3B-4bit", + "35b-mlx-4bit": "mlx-community/Qwen3.6-35B-A3B-4bit", + "35b-mlx-6bit": "mlx-community/Qwen3.6-35B-A3B-6bit", + "35b-mlx-8bit": "mlx-community/Qwen3.6-35B-A3B-8bit" + }, + "default": "35b-mlx", + "platform_defaults": { + "darwin/arm64": "35b-mlx", + "linux": "27b-q4_k_m", + "windows": "27b-q4_k_m" + }, + "sizes": { + "27b-q4_k_m": "~19.1 GiB", + "27b-q8_0": "~28.6 GiB", + "35b-mlx": "~20 GiB", + "35b-mlx-4bit": "~20 GiB", + "35b-mlx-6bit": "~27 GiB", + "35b-mlx-8bit": "~36 GiB" + }, + "memory": { + "27b-q4_k_m": "24 GiB min; 32 GiB recommended", + "27b-q8_0": "40 GiB min; 48 GiB recommended" + }, + "gpus": { + "27b-q4_k_m": "Optional Metal, CUDA, Vulkan, or ROCm acceleration", + "27b-q8_0": "Optional Metal, CUDA, Vulkan, or ROCm acceleration" + }, + "features": { + "27b-q4_k_m": "reasoning, coding, tool use, image understanding with optional mmproj", + "27b-q8_0": "higher-fidelity reasoning, coding, tool use, image understanding with optional mmproj" + } + }, + "qwen3.8": { + "name": "qwen3.8", + "repo": "unsloth/Qwen3.8-27B-GGUF", + "files": { + "27b-mlx": "", + "27b-q4_k_m": "Qwen3.8-27B-UD-Q4_K_M.gguf", + "27b-q5_k_m": "Qwen3.8-27B-UD-Q5_K_M.gguf" + }, + "kind": "text", + "backends": { + "27b-mlx": "mlx-vlm" + }, + "repos": { + "27b-mlx": "mlx-community/Qwen3.8-27B-4bit" + }, + "default": "27b-q4_k_m", + "platform_defaults": { + "darwin/arm64": "27b-mlx" + }, + "sizes": { + "27b-mlx": "~15 GiB", + "27b-q4_k_m": "~16 GiB", + "27b-q5_k_m": "~19 GiB" + }, + "memory": { + "27b-mlx": "24 GiB min; 32 GiB recommended", + "27b-q4_k_m": "24 GiB min; 32 GiB recommended", + "27b-q5_k_m": "32 GiB min; 40 GiB recommended" + }, + "gpus": { + "27b-mlx": "Apple Silicon GPU", + "27b-q4_k_m": "Optional GPU acceleration", + "27b-q5_k_m": "Optional GPU acceleration" + }, + "features": { + "27b-mlx": "agentic coding, tool use, Apple Silicon optimized", + "27b-q4_k_m": "agentic coding, tool use, long-horizon reasoning", + "27b-q5_k_m": "agentic coding, tool use, long-horizon reasoning" + } + }, + "sd-turbo": { + "name": "sd-turbo", + "repo": "stabilityai/sd-turbo", + "files": { + "fp16": "", + "onnx-arm64": "", + "onnx-directml": "" + }, + "kind": "image", + "backends": { + "fp16": "diffusers", + "onnx-arm64": "onnx-cpu", + "onnx-directml": "onnx-directml" + }, + "repos": { + "onnx-arm64": "Heliosoph/sd-turbo-onnx", + "onnx-directml": "Heliosoph/sd-turbo-onnx" + }, + "default": "fp16", + "width": 512, + "height": 512, + "steps": 4, + "sizes": { + "fp16": "~3 GiB", + "onnx-arm64": "~4.8 GiB", + "onnx-directml": "~4.8 GiB" + }, + "memory": { + "fp16": "16 GiB min; 24 GiB recommended", + "onnx-arm64": "12 GiB min; 16 GiB recommended", + "onnx-directml": "12 GiB min; 16 GiB recommended" + }, + "gpus": { + "fp16": "NVIDIA CUDA, 6 GiB+ VRAM", + "onnx-arm64": "CPU (native Windows ARM64)", + "onnx-directml": "AMD, Intel, or NVIDIA DirectX 12 GPU" + } + }, + "sdxl-turbo": { + "name": "sdxl-turbo", + "repo": "stabilityai/sdxl-turbo", + "files": { + "fp16": "" + }, + "kind": "image", + "backends": { + "fp16": "diffusers" + }, + "default": "fp16", + "width": 1024, + "height": 1024, + "steps": 4, + "sizes": { + "fp16": "~7 GiB" + }, + "memory": { + "fp16": "24 GiB min; 32 GiB recommended" + }, + "gpus": { + "fp16": "NVIDIA CUDA, 8 GiB+ VRAM" + } + }, + "spark-x2.5-4b": { + "context": 8192, + "name": "spark-x2.5-4b", + "repo": "XHToken/Spark-X2.5-4B-GGUF", + "files": { + "q4_k_m": "Spark-X2.5-4B-Q4_K_M.gguf" + }, + "default": "q4_k_m", + "sizes": { + "q4_k_m": "~2.42 GiB" + }, + "memory": { + "q4_k_m": "8 GiB test target; 16 GiB recommended at 8K context" + }, + "gpus": { + "q4_k_m": "CPU or Apple Metal; Vulkan pending qualification" + }, + "languages": { + "q4_k_m": "multilingual" + }, + "features": { + "q4_k_m": "basic text chat; tools and long context not yet qualified" + }, + "downloads": { + "q4_k_m": { + "revision": "902d865994943ab9235670e24f01846ee06091f2", + "sha256": "adfcfa19a4ed6a5985da8bf565fe15f8e1a7e131d79bae2d19d48d1c40109428", + "size_bytes": 2600224352 + } + }, + "license": "Apache-2.0", + "license_url": "https://huggingface.co/XHToken/Spark-X2.5-4B" }, "stable-audio-3": { "name": "stable-audio-3", @@ -1255,6 +1215,84 @@ "gated": true, "license": "Stability AI Community License", "license_url": "https://huggingface.co/collections/stabilityai/stable-audio-3" + }, + "stable-video-diffusion": { + "name": "stable-video-diffusion", + "repo": "stabilityai/stable-video-diffusion-img2vid-xt", + "files": { + "xt-fp16": "" + }, + "kind": "video", + "backends": { + "xt-fp16": "diffusers-video" + }, + "default": "xt-fp16", + "width": 1024, + "height": 576, + "steps": 25, + "frames": 25, + "fps": 7, + "sizes": { + "xt-fp16": "~4.5 GiB" + }, + "memory": { + "xt-fp16": "16 GiB min; 24 GiB recommended" + }, + "gpus": { + "xt-fp16": "NVIDIA CUDA, 8 GiB+ VRAM; CPU offload" + } + }, + "wan2.2-video": { + "name": "wan2.2-video", + "repo": "Anes1032/Wan2.2-TI2V-5B-mlx-q8", + "files": { + "5b-q8-mlx": "" + }, + "kind": "video", + "backends": { + "5b-q8-mlx": "mlx-video" + }, + "default": "5b-q8-mlx", + "width": 832, + "height": 480, + "steps": 40, + "frames": 41, + "fps": 24, + "sizes": { + "5b-q8-mlx": "~18 GiB" + }, + "memory": { + "5b-q8-mlx": "32 GiB min; 48 GiB recommended" + }, + "gpus": { + "5b-q8-mlx": "Apple Silicon GPU" + } + }, + "yume-video": { + "name": "yume-video", + "repo": "ckurasek/Yume-1.5-5B-720P-MLX", + "files": { + "5b-mlx": "" + }, + "kind": "video", + "backends": { + "5b-mlx": "mlx-video" + }, + "default": "5b-mlx", + "width": 832, + "height": 480, + "steps": 30, + "frames": 81, + "fps": 24, + "sizes": { + "5b-mlx": "~23 GiB" + }, + "memory": { + "5b-mlx": "32 GiB min; 48 GiB recommended" + }, + "gpus": { + "5b-mlx": "Apple Silicon GPU" + } } } } diff --git a/catalog/catalog.json.sha256 b/catalog/catalog.json.sha256 index 494ec75..5b8400e 100644 --- a/catalog/catalog.json.sha256 +++ b/catalog/catalog.json.sha256 @@ -1 +1 @@ -b408b5e1cd22626c60dfa56d49ac6543b4a699af3ce4ca96b48cf8869cdf3c9f catalog.json +afdff47ef1e9628ac008abe8a2372c3553a8819aa4be29b7272ea23fe137fc6e catalog.json diff --git a/desktop/scripts/prepare-runtime.ts b/desktop/scripts/prepare-runtime.ts index 39e0ffb..4c08c5e 100644 --- a/desktop/scripts/prepare-runtime.ts +++ b/desktop/scripts/prepare-runtime.ts @@ -7,14 +7,14 @@ import path from "node:path"; import { fileURLToPath } from "node:url"; import { verifyBundledRuntime } from "./verify-runtime.ts"; -export const LLAMA_CPP_VERSION = "b10964"; +export const LLAMA_CPP_VERSION = "b11476"; const runtimeChecksums: Record = { - "darwin/arm64": "033c845c1df9bf945ff37bb193238b40910b2244be3e1e637b2ceb5878f1a6f5", - "win32/x64": "1ee3ad952f4ba71f438bd6d7bebef19e1c7af04adcaa35d08b4ddabb27d4c642", - "win32/arm64": "4b6a004b076eea47c318bea35cf1db2ff2bf037738b04645646ae8d7c3159478", - "linux/x64": "55d1e58e14c11eedea090bf088fdeefbfe7b4b09ee03bf6dba9834651769afcf", - "linux/arm64": "f7864baa0edf5a059fb42c5efb5aceb96075aa1f41e6c3142b71ca69286cb0bb", + "darwin/arm64": "577634a1b8a59e8dabe02ba10de1e610be0574dfaf1cf3020e6dd42853ed877e", + "win32/x64": "5c71e7b749697da4a8d46e9ee55486845cbba27c9dfbecb4007f31ba6610d523", + "win32/arm64": "68e3a218ed7d9cd563e8ddf7a1e58d88d034a8f90a91061bdd3876bf247d8a93", + "linux/x64": "5bb4306d7917f33e81efda02e6f791ae6a82e86bee121227a3ab2b4e8e40427f", + "linux/arm64": "8dae2f39afee01d3032a101d7c398a6734d6fd9697690d8f93bdce4e3a9efea2", }; export function windowsExtraction(archive: string, destination: string) { diff --git a/desktop/src/app/adapters.ts b/desktop/src/app/adapters.ts index 36b1300..abf9b33 100644 --- a/desktop/src/app/adapters.ts +++ b/desktop/src/app/adapters.ts @@ -95,6 +95,7 @@ export function createRendererAdapters( installed: Boolean(local), installedBytes: local ? requirements.diskBytes : undefined, gated: model.gated, + licenseAcceptanceRequired: model.license_acceptance_required, license: model.license, licenseUrl: model.license_url, }; @@ -118,7 +119,7 @@ export function createRendererAdapters( name: modelId, jobId, acceptLicense: options.acceptLicense, - accessToken: options.accessToken, + accessToken: options.accessToken?.trim() || undefined, }); const refreshed = await listModels(); const installed = refreshed.find((model) => model.id === modelId); diff --git a/desktop/src/features/models/ModelsScreen.tsx b/desktop/src/features/models/ModelsScreen.tsx index b2bb706..637ac6c 100644 --- a/desktop/src/features/models/ModelsScreen.tsx +++ b/desktop/src/features/models/ModelsScreen.tsx @@ -361,7 +361,7 @@ export function ModelsScreen({ adapter, machine }: ModelsScreenProps) { transfer={transfers[model.id]} onCancel={() => void cancelPull(model)} onOpen={() => setSelected(model)} - onPull={() => model.gated ? setSelected(model) : void pull(model)} + onPull={() => model.gated || model.licenseAcceptanceRequired ? setSelected(model) : void pull(model)} /> ))} @@ -610,14 +610,13 @@ function ModelDetail({ {tag} ))} - {model.gated && !model.installed ? ( + {(model.gated || model.licenseAcceptanceRequired) && !model.installed ? (
- {model.license ?? "Gated model license"} + {model.license ?? "Restricted model license"}

- First accept the provider terms at {model.licenseUrl}. Paste a Hugging Face read - token for this download. Tapioca passes it only to the local downloader and does not save it. + Review the provider terms at {model.licenseUrl} before downloading.

- : null}