From c0e132d973276e5c1213bd03d930718ff056fd57 Mon Sep 17 00:00:00 2001 From: Hongsheng Liu Date: Mon, 18 May 2026 00:02:34 +0800 Subject: [PATCH] [Doc] Reorganize available recipes into a table (#3671) Signed-off-by: hsliu Co-authored-by: deepseek-v4-pro --- README.md | 2 +- recipes/Baidu/ERNIE-Image.md | 4 +- recipes/LTX/LTX-2.3.md | 2 +- recipes/LTX/LTX-2.md | 4 +- recipes/Qwen/Qwen-Image.md | 4 +- recipes/Qwen/Qwen3-Omni.md | 2 + recipes/Qwen/Qwen3-TTS.md | 4 +- recipes/README.md | 42 ++++++++----------- recipes/SenseNova/SenseNova-U1.md | 4 +- recipes/Tencent/Covo-Audio-Chat.md | 4 +- .../HunyuanImage-3.0-Instruct.md | 5 +-- recipes/Wan-AI/Wan2.2-I2V.md | 4 +- recipes/Wan-AI/Wan2.2-S2V.md | 4 +- recipes/fishaudio/Fish-Speech-S2-Pro.md | 2 + recipes/inclusionAI/Ming-flash-omni-2.0.md | 4 +- 15 files changed, 52 insertions(+), 39 deletions(-) rename recipes/{Tencent-Hunyuan => Tencent}/HunyuanImage-3.0-Instruct.md (97%) diff --git a/README.md b/README.md index f4a05fa5f..7dfc9cb01 100644 --- a/README.md +++ b/README.md @@ -9,7 +9,7 @@ Easy, fast, and cheap omni-modality model serving for everyone

-| Documentation | User Forum | Developer Slack | WeChat | Paper | Slides | +| Documentation | DeepWiki | User Forum | Developer Slack | WeChat | Paper | Slides |

diff --git a/recipes/Baidu/ERNIE-Image.md b/recipes/Baidu/ERNIE-Image.md index b09b61c9c..ccef7807e 100644 --- a/recipes/Baidu/ERNIE-Image.md +++ b/recipes/Baidu/ERNIE-Image.md @@ -1,4 +1,6 @@ -# ERNIE-Image for text-to-image generation +# ERNIE-Image + +> Text-to-image online serving (ERNIE-Image 8B) ## Summary diff --git a/recipes/LTX/LTX-2.3.md b/recipes/LTX/LTX-2.3.md index 8d92562fb..36a9eac09 100644 --- a/recipes/LTX/LTX-2.3.md +++ b/recipes/LTX/LTX-2.3.md @@ -1,4 +1,4 @@ -# LTX-2.3 Text-to-Video with Audio on 1x GPU (96GB VRAM) +# LTX-2.3 > 22B parameter text-to-video + audio generation model served via vLLM-Omni diff --git a/recipes/LTX/LTX-2.md b/recipes/LTX/LTX-2.md index ad5482f7a..df049f5c6 100644 --- a/recipes/LTX/LTX-2.md +++ b/recipes/LTX/LTX-2.md @@ -1,6 +1,6 @@ -# LTX-2 T2V and I2V +# LTX-2 -> LTX-2 for text-to-video & image-to-video on 1x H200 141GB +> Text-to-video and image-to-video serving ## Summary diff --git a/recipes/Qwen/Qwen-Image.md b/recipes/Qwen/Qwen-Image.md index de0d73c07..d9d60bf0d 100644 --- a/recipes/Qwen/Qwen-Image.md +++ b/recipes/Qwen/Qwen-Image.md @@ -1,4 +1,6 @@ -# Qwen-Image for text-to-image serving with optional continuous batching on 1x A100 80GB +# Qwen-Image + +> Text-to-image serving with optional step-wise continuous batching ## Summary diff --git a/recipes/Qwen/Qwen3-Omni.md b/recipes/Qwen/Qwen3-Omni.md index 49552d871..0ec67158c 100644 --- a/recipes/Qwen/Qwen3-Omni.md +++ b/recipes/Qwen/Qwen3-Omni.md @@ -1,5 +1,7 @@ # Qwen3-Omni +> Online serving for multimodal chat + ## Summary - Vendor: Qwen diff --git a/recipes/Qwen/Qwen3-TTS.md b/recipes/Qwen/Qwen3-TTS.md index b42fdf630..73a14a630 100644 --- a/recipes/Qwen/Qwen3-TTS.md +++ b/recipes/Qwen/Qwen3-TTS.md @@ -1,4 +1,6 @@ -# Qwen3-TTS for text-to-speech +# Qwen3-TTS + +> Text-to-speech serving (CustomVoice / VoiceDesign / Base) ## Summary diff --git a/recipes/README.md b/recipes/README.md index bccc4b222..09b1aeaee 100644 --- a/recipes/README.md +++ b/recipes/README.md @@ -19,34 +19,28 @@ recipes/ Qwen/ Qwen3-Omni.md Qwen3-TTS.md - Tencent-Hunyuan/ - HunyuanVideo.md + Tencent/ + Covo-Audio-Chat.md ``` ## Available Recipes -- [`Qwen/Qwen-Image.md`](./Qwen/Qwen-Image.md): text-to-image serving recipe for - Qwen-Image on `1x A100 80GB`, including optional step-wise continuous batching replay -- [`Qwen/Qwen3-Omni.md`](./Qwen/Qwen3-Omni.md): online serving recipe for - multimodal chat on `1x A100 80GB` -- [`Tencent/Covo-Audio-Chat.md`](./Tencent/Covo-Audio-Chat.md): online - serving recipe for audio chat on `1x A100 80GB` -- [`Qwen/Qwen3-TTS.md`](./Qwen/Qwen3-TTS.md): text-to-speech serving recipe - for Qwen3-TTS (CustomVoice / VoiceDesign / Base) on `1x H100/A100 80GB` -- [`LTX/LTX-2.md`](./LTX/LTX-2.md): text-to-video and image-to-video serving - recipe for LTX-2 on `1x H200 141GB` -- [`Wan-AI/Wan2.2-I2V.md`](./Wan-AI/Wan2.2-I2V.md): image-to-video serving - recipe for Wan2.2 14B on `8x Ascend NPU (A2/A3)` -- [`Tencent-Hunyuan/HunyuanImage-3.0-Instruct.md`](./Tencent-Hunyuan/HunyuanImage-3.0-Instruct.md): - DiT-only text-to-image serving and benchmark recipe for HunyuanImage-3.0-Instruct - on `4x H100/H800 80GB` -- [`inclusionAI/Ming-flash-omni-2.0.md`](./inclusionAI/Ming-flash-omni-2.0.md): - online serving recipe for multimodal chat (`4x H100 80GB`) and standalone TTS (`1x H100 80GB`) -- [`Baidu/ERNIE-Image.md`](./Baidu/ERNIE-Image.md): text-to-image serving - online serving recipe for ERNIE-Image 8B on `1x RTX 4090 24GB` or `2x RTX 4090 24GB` -- [`fishaudio/Fish-Speech-S2-Pro.md`](./fishaudio/Fish-Speech-S2-Pro.md): online serving recipe for TTS on `1x A800 80GB` -- [`audiox/AudioX.md`](./audiox/AudioX.md): offline + online recipe for AudioX - unified text/video→audio diffusion on `1x L4 24GB` +| Recipe | Task | Hardware | +|--------|------|----------| +| [`audiox/AudioX.md`](./audiox/AudioX.md) | Offline + online unified text/video→audio diffusion | 1x L4 24GB | +| [`Baidu/ERNIE-Image.md`](./Baidu/ERNIE-Image.md) | Text-to-image online serving (ERNIE-Image 8B) | 1x or 2x RTX 4090 24GB | +| [`fishaudio/Fish-Speech-S2-Pro.md`](./fishaudio/Fish-Speech-S2-Pro.md) | Online serving for TTS | 1x A800 80GB | +| [`inclusionAI/Ming-flash-omni-2.0.md`](./inclusionAI/Ming-flash-omni-2.0.md) | Online serving for multimodal chat + standalone TTS | 4x H100 / 1x H100 80GB | +| [`LTX/LTX-2.md`](./LTX/LTX-2.md) | Text-to-video and image-to-video serving | 1x H200 141GB | +| [`LTX/LTX-2.3.md`](./LTX/LTX-2.3.md) | Text-to-video with audio generation (22B) | 1x GPU (96GB VRAM) | +| [`Qwen/Qwen-Image.md`](./Qwen/Qwen-Image.md) | Text-to-image serving with step-wise continuous batching replay | 1x A100 80GB | +| [`Qwen/Qwen3-Omni.md`](./Qwen/Qwen3-Omni.md) | Online serving for multimodal chat | 1x A100 80GB | +| [`Qwen/Qwen3-TTS.md`](./Qwen/Qwen3-TTS.md) | Text-to-speech serving (CustomVoice / VoiceDesign / Base) | 1x H100/A100 80GB | +| [`SenseNova/SenseNova-U1.md`](./SenseNova/SenseNova-U1.md) | Unified image generation and understanding | 1x H200 (144GB) | +| [`Tencent/Covo-Audio-Chat.md`](./Tencent/Covo-Audio-Chat.md) | Online serving for audio chat | 1x A100 80GB | +| [`Tencent/HunyuanImage-3.0-Instruct.md`](./Tencent/HunyuanImage-3.0-Instruct.md) | DiT-only text-to-image serving and benchmark | 4x H100/H800 80GB | +| [`Wan-AI/Wan2.2-I2V.md`](./Wan-AI/Wan2.2-I2V.md) | Image-to-video serving (Wan2.2 14B) | 8x Ascend NPU (A2/A3) | +| [`Wan-AI/Wan2.2-S2V.md`](./Wan-AI/Wan2.2-S2V.md) | Speech-to-video serving (Wan2.2 14B) | 2x A100/H100 80GB | Within a single recipe file, include different hardware support sections such as `GPU`, `ROCm`, and `NPU`, and add concrete tested configurations like diff --git a/recipes/SenseNova/SenseNova-U1.md b/recipes/SenseNova/SenseNova-U1.md index e88da2eb3..b1a14eaa2 100644 --- a/recipes/SenseNova/SenseNova-U1.md +++ b/recipes/SenseNova/SenseNova-U1.md @@ -1,4 +1,6 @@ -# SenseNova-U1 for unified image generation and understanding +# SenseNova-U1 + +> Unified image generation and understanding ## Summary diff --git a/recipes/Tencent/Covo-Audio-Chat.md b/recipes/Tencent/Covo-Audio-Chat.md index 4f2eab635..e1e29149a 100644 --- a/recipes/Tencent/Covo-Audio-Chat.md +++ b/recipes/Tencent/Covo-Audio-Chat.md @@ -1,4 +1,6 @@ -# Covo-Audio-Chat for audio chat on 1x A100 80GB +# Covo-Audio-Chat + +> Online serving for audio chat ## Summary diff --git a/recipes/Tencent-Hunyuan/HunyuanImage-3.0-Instruct.md b/recipes/Tencent/HunyuanImage-3.0-Instruct.md similarity index 97% rename from recipes/Tencent-Hunyuan/HunyuanImage-3.0-Instruct.md rename to recipes/Tencent/HunyuanImage-3.0-Instruct.md index 1694cc241..9ccb37bed 100644 --- a/recipes/Tencent-Hunyuan/HunyuanImage-3.0-Instruct.md +++ b/recipes/Tencent/HunyuanImage-3.0-Instruct.md @@ -1,7 +1,6 @@ -# HunyuanImage-3.0-Instruct DiT Image Generation on 4x GPU +# HunyuanImage-3.0-Instruct -> DiT-only text-to-image recipe for HunyuanImage-3.0-Instruct with FP8, -> tensor parallelism, sequence parallelism, and CFG parallelism. +> DiT-only text-to-image serving and benchmark ## Summary diff --git a/recipes/Wan-AI/Wan2.2-I2V.md b/recipes/Wan-AI/Wan2.2-I2V.md index faa6a9e2a..5c5dd984b 100644 --- a/recipes/Wan-AI/Wan2.2-I2V.md +++ b/recipes/Wan-AI/Wan2.2-I2V.md @@ -1,4 +1,6 @@ -# Wan2.2 Image To Video +# Wan2.2 Image-to-Video + +> Image-to-video serving (Wan2.2 14B) ## Summary diff --git a/recipes/Wan-AI/Wan2.2-S2V.md b/recipes/Wan-AI/Wan2.2-S2V.md index 34470ed0a..9f34509d4 100644 --- a/recipes/Wan-AI/Wan2.2-S2V.md +++ b/recipes/Wan-AI/Wan2.2-S2V.md @@ -1,4 +1,6 @@ -# Wan2.2 Speech To Video +# Wan2.2 Speech-to-Video + +> Speech-to-video serving (Wan2.2 14B) ## Summary diff --git a/recipes/fishaudio/Fish-Speech-S2-Pro.md b/recipes/fishaudio/Fish-Speech-S2-Pro.md index 20cfdce61..91ac3fe03 100644 --- a/recipes/fishaudio/Fish-Speech-S2-Pro.md +++ b/recipes/fishaudio/Fish-Speech-S2-Pro.md @@ -1,5 +1,7 @@ # Fish Speech S2 Pro +> Online serving for TTS + ## Summary - Vendor: FishAudio diff --git a/recipes/inclusionAI/Ming-flash-omni-2.0.md b/recipes/inclusionAI/Ming-flash-omni-2.0.md index 71a90cbe7..3dcda2cbe 100644 --- a/recipes/inclusionAI/Ming-flash-omni-2.0.md +++ b/recipes/inclusionAI/Ming-flash-omni-2.0.md @@ -1,4 +1,6 @@ -# Ming-flash-omni 2.0 for omni-speech chat and standalone TTS +# Ming-flash-omni 2.0 + +> Online serving for multimodal chat + standalone TTS ## Summary