README_EN.md - VideoClaw

README_EN.md

1	<p align="center">
2	<img src="video-claw-pics/banner.png" width="100%" />
3	</p>
4
5	<h1 align="center">
6	VideoClaw: AI Creative Video Production Agent
7	</h1>
8
9	<p align="center">
10	<a href="./README.md">简体中文</a> \| <b>English</b>
11	</p>
12
13	<h3 align="center">
14	<img src="https://img.shields.io/badge/Version-1.0.0-blue.svg" alt="Version">
15	<a href="https://github.com/HITsz-TMG/VideoClaw/blob/main/LICENSE">
16	<img src="https://img.shields.io/github/license/HITsz-TMG/VideoClaw?style=flat-square" alt="License">
17	</a>
18	<a href="https://github.com/HITsz-TMG/VideoClaw/stargazers">
19	<img src="https://img.shields.io/github/stars/HITsz-TMG/VideoClaw?style=flat-square&logo=github" alt="Stars">
20	</a>
21	<a href="https://github.com/HITsz-TMG/VideoClaw/fork">
22	<img src="https://img.shields.io/github/forks/HITsz-TMG/VideoClaw?style=flat-square&logo=github" alt="Forks">
23	</a>
24	<img src="https://img.shields.io/badge/Python-3.9+-purple.svg" alt="Python">
25	<a href="#method-3-openclaw-auto-setup">
26	<img src="https://img.shields.io/badge/OpenClaw-Compatible-ff4444?logo=data:image/svg+xml;base64,PHN2ZyB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciIHZpZXdCb3g9IjAgMCAyNCAyNCI+PHBhdGggZD0iTTEyIDJDNi40OCAyIDIgNi40OCAyIDEyczQuNDggMTAgMTAgMTAgMTAtNC40OCAxMC0xMFMxNy41MiAyIDEyIDJ6IiBmaWxsPSJ3aGl0ZSIvPjwvc3ZnPg==" alt="OpenClaw Compatible">
27	</a>
28	</h3>
29
30	<p align="center">
31	<b><i><font size="5">Talk to <a href="https://github.com/openclaw/openclaw">OpenClaw</a> directly: "Generate a video about X" -> done.</font></i></b>
32	</p>
33
34	<div align="center">
35
36	📺 [Bilibili](https://space.bilibili.com/2031891503?spm_id_from=333.1007.0.0) ▶️ [YouTube](https://www.youtube.com/@imryanxu) 📖 [Integration Guide](#method-3-openclaw-auto-setup) 🦀 [ClawHub](https://clawhub.ai/hit-cxf/video-claw)
37
38	<a href="https://trendshift.io/repositories/24295" target="_blank"><img src="https://trendshift.io/api/badge/repositories/24295" alt="HITsz-TMG%2FVideo-Claw \| Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
39
40	</div>
41
42
43	# 💥 News
44
45	- `2026/3/27`: 🎬 VideoClaw was officially released, supporting full workflow automation from idea to video generation, with user intervention and adjustment available at any time.
46	- `2026/4/9`: ♾️ VideoClaw was optimized for short dramas, adding infinite continuation and customizable plots.
47	- `2026/4/29`: 🧩 Added three features: commentary-style short videos, action transfer, and digital human talking videos.
48	- `2026/5/8`: ⚙️ Added WebUI configuration for APIs and default models, with one-click installation support.
49	- `2026/5/13`: 🎞️ Integrated Pixelle-Video HTML templates into commentary-style short videos.
50
51
52	# 📖 Overview
53
54	<p align="center">
55	<img src="video-claw-pics/workflow.png" width="100%" />
56	</p>
57
58	VideoClaw is an AI director system for creative video production. You only need to provide an idea, a story outline, or even a vague concept. The system will break it down into an executable filmmaking workflow, continuously producing intermediate assets that can be reviewed, confirmed, revised, and delivered, until a complete final video is generated.
59
60	It is not a one-shot text-to-video tool. It is a full production line covering script planning -> character and scene design -> storyboard planning -> reference image generation -> video generation -> post-production editing. Instead of giving you a black-box result, VideoClaw behaves like a collaborative AI directing team: each stage informs the next, and every key node is visible, editable, and extensible.
61
62	# 📺 Demos
63
64	## 🎬 VideoClaw
65
66	<details>
67	<summary><b>Click to view WebUI UI design</b></summary>
68
69	\| Stage \| Demo \| Description \|
70	\|---\|---\|---\|
71	\| Home \| <img src="video-claw-pics/workflow_demo/homepage.png" width="800" /> \| Displays the system overview, supports viewing historical projects, creating new projects, and global configuration (API keys and default model settings). This is the starting point of the creation workflow. \|
72	\| Script Planning \| <img src="video-claw-pics/workflow_demo/stage-1.png" width="800" /> \| Enter a creative title and project synopsis, and the system automatically generates a structured multi-scene script (including narration and dialogue), with support for intelligent continuation of subsequent plots. \|
73	\| Character/Scene Design \| <img src="video-claw-pics/workflow_demo/stage-2.png" width="800" /> \| Automatically extracts core characteristics of characters and scenes from the script, generating stylistically consistent reference concept art as the visual basis for later storyboard generation. \|
74	\| Storyboard Planning \| <img src="video-claw-pics/workflow_demo/stage-3.png" width="800" /> \| Breaks each script scene into continuous visual storyboards, specifying camera perspective, action descriptions, and reference content in detail to ensure narrative coherence. \|
75	\| Reference Image Generation \| <img src="video-claw-pics/workflow_demo/stage-4.png" width="800" /> \| Generates high-quality, high-precision reference base images for each storyboard scene, controlling lighting details and composition as key visual references for video generation. \|
76	\| Video Generation \| <img src="video-claw-pics/workflow_demo/stage-5.png" width="800" /> \| Calls mainstream high-performance video generation models (such as Wan and Kling) to convert storyboard images into dynamic clips. \|
77	\| Final Editing \| <img src="video-claw-pics/workflow_demo/stage-6.png" width="800" /> \| Aggregates all generated video clips and exports a publishable final video with one click. \|
78
79	</details>
80
81	### 📱 Series 1: A Programmer Uses OpenClaw to Acquire His Former Company After Being Laid Off (Realistic Short Drama)
82	> 8 episodes in total, an underdog story with twists and reversal. The first 6 episodes were generated initially, followed by 2 continued episodes.
83
84	<table>
85	<tr>
86	<td align="center" valign="top" width="25%">
87	<a href="https://github.com/user-attachments/assets/1d095b82-3a72-4acc-9ca1-3ff4a6189232">
88	<img src="https://github.com/user-attachments/assets/47b1621e-5c5a-4cb1-9c51-dcf723ac5fda" width="100%" alt="Play Episode 1">
89	</a>
90	<br><b>▶️ Episode 1</b><br>Laid Off
91	</td>
92	<td align="center" valign="top" width="25%">
93	<a href="https://github.com/user-attachments/assets/489c5343-6345-4bce-81dc-bf6012b9c1cf">
94	<img src="https://github.com/user-attachments/assets/b0dd3781-b767-45c2-9fe2-3dc7421fbf80" width="100%" alt="Play Episode 2">
95	</a>
96	<br><b>▶️ Episode 2</b><br>Late-Night Departure, First Breakthrough
97	</td>
98	<td align="center" valign="top" width="25%">
99	<a href="https://github.com/user-attachments/assets/359809cf-678b-429c-bafa-55ff50fd3277">
100	<img src="https://github.com/user-attachments/assets/fcfd191c-61f9-484f-a036-ce717620c827" width="100%" alt="Play Episode 3">
101	</a>
102	<br><b>▶️ Episode 3</b><br>AI Funding, Old Employer in Crisis
103	</td>
104	<td align="center" valign="top" width="25%">
105	<a href="https://github.com/user-attachments/assets/5561a04a-5ab3-4099-bc63-2fdbcc48f8e9">
106	<img src="https://github.com/user-attachments/assets/14a535da-b0eb-4b4b-8039-517bac692696" width="100%" alt="Play Episode 4">
107	</a>
108	<br><b>▶️ Episode 4</b><br>Acquiring Xingyao
109	</td>
110	</tr>
111	<tr>
112	<td align="center" valign="top">
113	<a href="https://github.com/user-attachments/assets/ae4c3618-1990-4ff5-ad85-e09f23b08f7d">
114	<img src="https://github.com/user-attachments/assets/e2d7fbae-d945-461f-be2f-5eb781337cac" width="100%" alt="Play Episode 5">
115	</a>
116	<br><b>▶️ Episode 5</b><br>Acquisition, Liquidation, New Life
117	</td>
118	<td align="center" valign="top">
119	<a href="https://github.com/user-attachments/assets/4ddbb725-34d8-478b-97bc-5d7143f73101">
120	<img src="https://github.com/user-attachments/assets/02a8f7d4-04c9-460e-9ba1-07d594ffcd24" width="100%" alt="Play Episode 6">
121	</a>
122	<br><b>▶️ Episode 6</b><br>New Life, Looking Back
123	</td>
124	<td align="center" valign="top">
125	<a href="https://github.com/user-attachments/assets/1c1e5970-aaea-44ba-b041-5d551905bfde">
126	<img src="https://github.com/user-attachments/assets/053be420-20e8-423e-8379-f1d1554546c5" width="100%" alt="Play Episode 7">
127	</a>
128	<br><b>▶️ Episode 7</b><br>Technology Backfires
129	</td>
130	<td align="center" valign="top">
131	<a href="https://github.com/user-attachments/assets/e56e6784-6e49-4891-b0bc-a32308dd2145">
132	<img src="https://github.com/user-attachments/assets/19a50fcd-0a5a-4649-bb63-c7867e945d46" width="100%" alt="Play Episode 8">
133	</a>
134	<br><b>▶️ Episode 8</b><br>Upholding Ethics, Getting Through Together
135	</td>
136	</tr>
137	</table>
138
139	<br>
140
141	### 🖥️ Series 2: The Village Teacher (Sci-Fi Comic Drama)
142	> 5 episodes in total, a tribute to the inheritance of civilization.
143
144	<table>
145	<tr>
146	<td align="center" valign="top" width="50%">
147	<a href="https://github.com/user-attachments/assets/1ffe7b06-73e9-44cd-ad3f-afced5239f97">
148	<img src="https://github.com/user-attachments/assets/e1328cf7-23fe-48d8-9ae6-a7beeae6dda9" width="100%" alt="Play Episode 1">
149	</a>
150	<br><b>▶️ Episode 1</b><br>The Last Lesson
151	</td>
152	<td align="center" valign="top" width="50%">
153	<a href="https://github.com/user-attachments/assets/7547e5d3-872c-4344-8727-ee2be109797d">
154	<img src="https://github.com/user-attachments/assets/65bc3866-6405-4033-97b2-54de42a61402" width="100%" alt="Play Episode 2">
155	</a>
156	<br><b>▶️ Episode 2</b><br>The Cleansing Plan
157	</td>
158	</tr>
159	<tr>
160	<td align="center" valign="top">
161	<a href="https://github.com/user-attachments/assets/affed408-4df7-4ce7-9681-8f4ed45a6fcf">
162	<img src="https://github.com/user-attachments/assets/81acd075-e4f1-4621-b2dc-b2d87ed83b81" width="100%" alt="Play Episode 3">
163	</a>
164	<br><b>▶️ Episode 3</b><br>A Dying Entrustment
165	</td>
166	<td align="center" valign="top">
167	<a href="https://github.com/user-attachments/assets/dc7a85a8-6912-4443-a995-3d8f3ca30bc8">
168	<img src="https://github.com/user-attachments/assets/738c539e-8f90-47f5-9983-f3ac35d3d385" width="100%" alt="Play Episode 4">
169	</a>
170	<br><b>▶️ Episode 4</b><br>Questions of Life and Death
171	</td>
172	</tr>
173	<tr>
174	<td align="center" valign="top">
175	<a href="https://github.com/user-attachments/assets/1fb889c5-0e2b-40fa-a438-7399322ada47">
176	<img src="https://github.com/user-attachments/assets/ce4a1ac4-7308-43ba-a803-fc77b9b9561e" width="100%" alt="Play Episode 5">
177	</a>
178	<br><b>▶️ Episode 5</b><br>Light of Civilization
179	</td>
180	<td align="center" valign="top">
181	<!-- Empty cell to keep table layout aligned -->
182	</td>
183	</tr>
184	</table>
185
186	<br>
187
188	### 🎞️ More Demos
189	<details>
190	<summary><b>More Micro-Drama Clips</b></summary>
191
192	<table>
193	<tr>
194	<td align="center" valign="top" width="33%">
195	<video src="https://github.com/user-attachments/assets/63c2f33c-da50-44f0-8c26-a65611479d6a" controls width="100%" preload="none"></video>
196	<br><b>London Mystery</b>
197	</td>
198	<td align="center" valign="top" width="33%">
199	<video src="https://github.com/user-attachments/assets/d7c65cad-05b9-46c8-ab0e-96e39909f978" controls width="100%" preload="none"></video>
200	<br><b>A Dog's Purpose</b>
201	</td>
202	<td align="center" valign="top" width="33%">
203	<video src="https://github.com/user-attachments/assets/ec67546e-2d3d-4b34-b1ad-7d860a9bc1aa" controls width="100%" preload="none"></video>
204	<br><b>Drone Delivers Lychees</b>
205	</td>
206	</tr>
207	</table>
208
209	</details>
210
211	<br>
212
213	<details>
214	<summary><b>WeChat Interaction</b></summary>
215	<div align="center">
216
217	\| \| \| \| \|
218	\|:---:\|:---:\|:---:\|:---:\|
219	\| ![WeChat 1](video-claw-pics/wechat_demo/wechat_1.jpg) \| ![WeChat 2](video-claw-pics/wechat_demo/wechat_2.jpg) \| ![WeChat 3](video-claw-pics/wechat_demo/wechat_3.jpg) \| ![WeChat 4](video-claw-pics/wechat_demo/wechat_4.jpg) \|
220
221	</div>
222	</details>
223
224	<br>
225
226	<details>
227	<summary><b>Feishu Interaction</b></summary>
228	<div align="center">
229
230	\| \| \| \| \|
231	\|:---:\|:---:\|:---:\|:---:\|
232	\| ![Feishu 1](video-claw-pics/feishu_demo/feishu_1.jpg) \| ![Feishu 2](video-claw-pics/feishu_demo/feishu_2.jpg) \| ![Feishu 3](video-claw-pics/feishu_demo/feishu_3.jpg) \| ![Feishu 4](video-claw-pics/feishu_demo/feishu_4.jpg) \|
233
234	</div>
235	</details>
236
237	<!--
238	## 🧩 Quick Creation
239
240	<details>
241	<summary><b>Click to view WebUI UI design</b></summary>
242
243	\| Pipeline \| Demo \| Frontend Entry \| Description \|
244	\|---\|---\|---\|---\|
245	\| Artistic Short Video \| <img src="video-claw-pics/pipeline_demo/standard.png" width="600" /> \| Sidebar: "Artistic Short Video" \| Supports both "Image Montage / Dynamic Video" and "Creative Inspiration / Full Script" modes. The system splits narration by periods, generates an image and voiceover for each segment, then either composes image-based clips or calls an image-to-video model for dynamic segments. Optional title and subtitles are supported. \|
246	\| Action Transfer \| <img src="video-claw-pics/pipeline_demo/action_transfer.png" width="600" /> \| Sidebar: "Action Transfer" \| Takes a reference image, an action video, and a prompt, then calls a video model with action-transfer capability to generate the result video. \|
247	\| Digital Human Talking Video \| <img src="video-claw-pics/pipeline_demo/digital_human.png" width="600" /> \| Sidebar: "Digital Human Talking Video" \| Takes a character image and talking script, generates sentence-level speech and digital human video clips, uses the previous clip's tail frame to continue multi-clip generation, and replaces the final video audio with the generated speech. \|
248
249	</details>
250
251	### Artistic Short Video
252
253	<div align="center">
254	<table align="center" border="0" cellspacing="0" cellpadding="0" style="border:none; border-collapse:collapse; margin:0 auto;">
255	<tr>
256	<td align="center" valign="top" width="25%" style="border:none;"></td>
257	<td align="center" valign="top" width="25%" style="border:none;">
258	<video src="https://github.com/user-attachments/assets/7a674bb7-6ee9-4b83-bfd3-0d880127b632" controls width="100%" preload="none"></video>
259	<br><b>▶️ Mountain and River Dreams</b>
260	</td>
261	<td align="center" valign="top" width="25%" style="border:none;">
262	<video src="https://github.com/user-attachments/assets/a62c8184-322d-4c06-b16f-19660766e816" controls width="100%" preload="none"></video>
263	<br><b>▶️ The Vastness of Life</b>
264	</td>
265	<td align="center" valign="top" width="25%" style="border:none;"></td>
266	</tr>
267	</table>
268	</div>
269
270	<br>
271	-->
272
273	# ✨ Features
274
275	\| Capability \| Description \|
276	\|---\|---\|
277	\| 🎬 End-to-end generation from idea to final cut \| Connects scripts, characters, storyboards, reference images, video clips, and post-production into one complete workflow, upgrading scattered generation abilities into a full video production pipeline. \|
278	\| 🖼️ Storyboard-driven controllable creation \| Uses structured scripts, storyboard planning, and reference image generation to make character consistency, shot expression, and visual style more stable and controllable. \|
279	\| ✍️ Editable, continuable, and regenerable \| Supports intelligent continuation of plots and storyboards, while also allowing character, reference image, and video stages to be edited and regenerated without starting from scratch. \|
280	\| 🧩 Lightweight Pipeline tasks \| Supports three one-shot tasks: commentary-style short videos, action transfer, and digital human talking videos. They are suitable for batch generation of image/text or dynamic short videos, action-transfer videos, and talking-head videos. \|
281	\| 📲 Local deployment, multi-platform collaboration, and asset retention \| Supports Web UI, WeChat / Feishu collaboration, OpenClaw Skill integration, and full-chain retention of scripts, images, video clips, and final outputs. \|
282
283	---
284
285	# 🚀 Quick Start
286
287	## Method 1: One-Click Installation (Recommended)
288
289	Linux / MacOS:
290
291	```bash
292	# 1. Clone the repository
293	git clone https://github.com/HITsz-TMG/VideoClaw.git
294	cd VideoClaw
295
296	# 2. Enter the app directory and run the installer
297	cd video-claw/video-claw
298	chmod +x install.sh
299	./install.sh
300
301	# 3. back to root dir
302	cd ../..
303	```
304
305	Windows:
306
307	```bat
308	# 1. Clone the repository
309	git clone https://github.com/HITsz-TMG/VideoClaw.git
310	cd VideoClaw
311
312	# 2. Enter the app directory and run the installer
313	cd video-claw\video-claw
314	install.bat
315
316	# 3. back to root dir
317	cd ../..
318	```
319
320	The installer checks Python, Node.js, npm, and ffmpeg, installs backend and frontend dependencies, copies `backend/config.yaml.example` to `backend/config.yaml`, and builds the frontend. After installation, fill in model service API keys in `backend/config.yaml` and confirm the main-workflow default models under `models`. You can also start the frontend and edit these settings from the "Settings" page at the bottom of the sidebar. Then start the services:
321
322	```bash
323	# Start backend
324	cd video-claw/video-claw/backend
325	uv run python api_server.py
326
327	# Start frontend in a new terminal
328	cd video-claw/video-claw/frontend
329	npm start
330	```
331
332	By default, the backend runs at `http://localhost:8000`, and the frontend runs at `http://localhost:3000`.
333
334	If you only want to install dependencies and skip the frontend build temporarily, run:
335
336	```bash
337	AIGC_DIRECTOR_SKIP_FRONTEND_BUILD=1 ./install.sh
338	```
339
340	## Method 2: Manual Installation
341
342	```bash
343	# 1. Clone the repository
344	git clone https://github.com/HITsz-TMG/Video-Claw.git
345	cd Video-Claw
346
347	# 2. Configure and start the backend
348	cd video-claw/video-claw/backend
349
350	# Install backend dependencies
351	uv sync
352
353	# Configure backend YAML
354	cp config.yaml.example config.yaml
355	# Edit config.yaml, fill in API keys, and confirm main-workflow default models
356	# You can also use the frontend Settings page after startup
357
358	# Start backend
359	uv run python api_server.py
360	# Service runs at http://localhost:8000
361	```
362
363	```bash
364	# 3. Configure and start the frontend in a new terminal
365	cd video-claw/video-claw/frontend
366	npm install
367	npm run build
368	npm start
369	# Visit http://localhost:3000
370	```
371
372	If `uv` is not installed, you can also create a Python virtual environment manually and install backend dependencies with `pip install -r requirements.txt`.
373
374	## Method 3: OpenClaw Auto Setup
375
376	Send this message to OpenClaw:
377
378	```text
379	Please clone this git repository: https://github.com/HITsz-TMG/Video-Claw.git
380	Then recursively copy the video-claw folder inside Video-Claw to .openclaw/workspace/skills and use it as an AIGC-related skill.
381	```
382
383	When using it, it is recommended to explicitly say "use video-claw":
384
385	```text
386	Use video-claw to generate a video with the content "A Dog's Purpose".
387	```
388
389	## Method 4: Install via ClawHub
390
391	Make sure `clawhub-cli` is installed locally.
392
393	Open a terminal and run the following command. Choose `yes` for every prompt.
394
395	```bash
396	clawhub install video-claw
397	```
398
399	After installation, ClawHub will copy `video-claw` into `workspace/skills` or your specified skills directory.
400
401	Then you can follow Method 1 for one-click installation or Method 2 for manual setup, or let OpenClaw build and run the project for you.
402
403	The first time you use `video-claw`, if the project has not been built manually, OpenClaw will automatically build and start both the backend and frontend. This may take some time because setup involves dependency installation and compilation.
404
405	---
406
407	# 🔧 Configuration
408
409	<details>
410	<summary><b>Click to expand full requirements and configuration</b></summary>
411
412	## Requirements
413
414	- Python: 3.9+
415	- Node.js: 18+
416	- npm: 9+
417
418	## Backend Configuration
419
420	Backend configuration is stored in `video-claw/backend/config.yaml` using a lowercase hierarchical YAML structure. You can edit this file directly, or open the frontend "Settings" page from the bottom of the sidebar.
421
422	- `api_providers` stores API keys, base URLs, and proxy toggles for each model provider.
423	- `models` stores default models for the main workflow home page. When creating a project, the frontend reads these defaults first and sends the concrete model parameters to the backend. The backend no longer silently chooses fallback models for the main workflow; missing model parameters will return an error.
424	- Pipelines (Artistic Short Video, Action Transfer, and Digital Human Talking Video) do not use these main-workflow defaults. Choose their models separately on each Pipeline page.
425
426	## Frontend Settings Page
427
428	After starting both the backend and frontend, open the "Settings" page from the bottom of the left sidebar. You do not have to edit YAML manually for common configuration:
429
430	- Fill in or update API Key / Access Key / Secret Key values for OpenAI, Gemini, DeepSeek, DashScope, Volcengine ARK, Kling, and other configured providers.
431	- Edit each provider's `base_url`, `enable_proxy`, and the shared proxy address `api_providers.common.proxy`.
432	- Select main-workflow default models, including `llm`, `vlm`, `image_t2i`, `image_it2i`, `video`, `video_ratio`, and `eval`.
433	- Saving writes the values back to `backend/config.yaml`. API keys, proxy settings, and default models are read by newly created projects; service startup fields such as `server.host` and `server.port` require restarting the backend to fully take effect.
434
435	```yaml
436	project_name: Video-Claw
437
438	server:
439	host: 127.0.0.1
440	port: 8000
441	debug: false
442
443	api_providers:
444	common:
445	print_model_input: false
446	proxy: ''
447	openai:
448	api_key: your_openai_key
449	base_url: https://api.openai.com/v1
450	enable_proxy: false
451	gemini:
452	api_key: your_gemini_key
453	base_url: https://generativelanguage.googleapis.com/v1beta
454	enable_proxy: false
455	deepseek:
456	api_key: your_deepseek_key
457	base_url: https://api.deepseek.com/v1
458	enable_proxy: false
459	dashscope:
460	api_key: your_dashscope_key
461	base_url: https://dashscope.aliyuncs.com/api/v1
462	enable_proxy: false
463	ark:
464	api_key: your_ark_key
465	base_url: https://ark.cn-beijing.volces.com/api/v3
466	enable_proxy: false
467	kling:
468	access_key: your_kling_access_key
469	secret_key: your_kling_secret_key
470	enable_proxy: false
471
472	models:
473	llm: qwen3.5-plus
474	vlm: qwen3.5-plus
475	image_t2i: doubao-seedream-5-0-260128
476	image_it2i: doubao-seedream-5-0-260128
477	video: wan2.7-i2v
478	video_ratio: '16:9'
479	eval: qwen3.5-plus
480	```
481
482	<!-- `api_providers.common.proxy` is the only proxy address. Each provider decides whether to use it via `enable_proxy`, which is disabled by default, so different model calls in the same process do not pollute each other. Changes to service startup fields such as `server.host` and `server.port` require restarting the backend to fully take effect. API keys, proxy settings, and main-workflow defaults under `models` are read by new project creation and model calls. -->
483
484	## API Keys and Model Providers
485
486	\| Provider \| Config fields \| Common use \|
487	\|:---:\|:---\|:---\|
488	\| OpenAI \| `api_providers.openai.api_key` / `base_url` \| GPT text/vision models and OpenAI image models \|
489	\| Gemini \| `api_providers.gemini.api_key` / `base_url` \| Gemini text and vision models \|
490	\| DeepSeek \| `api_providers.deepseek.api_key` / `base_url` \| DeepSeek text models \|
491	\| DashScope \| `api_providers.dashscope.api_key` / `base_url` \| Qwen, Wan image/video models, and related Alibaba Cloud services \|
492	\| Volcengine ARK \| `api_providers.ark.api_key` / `base_url` \| Seedream image models and Seedance video models \|
493	\| Kling \| `api_providers.kling.access_key` / `secret_key` / `base_url` \| Kling video generation \|
494
495	You only need to fill in the provider keys required by the models you choose. For example, if the main workflow uses a `doubao-seedream-` image model, configure `ark.api_key`; if it uses a `wan` video model, configure `dashscope.api_key`. If you choose different models on a Pipeline page, make sure the corresponding provider key is also configured.
496
497	## Available Models
498
499	\| Type \| Models \|
500	\|:---:\|:---\|
501	\| LLM \| qwen3.6-max-preview, qwen3-max, deepseek-chat, deepseek-reasoner, deepseek-v4-flash, deepseek-v4-pro, gpt-4o, gpt-5, gpt-5.4, gemini-2.5-flash, gemini-2.0-flash, kimi-k2.6 \|
502	\| VLM \| qwen3.6-plus, qwen3.6-flash, kimi-k2.6, gpt-5.4, gemini-2.5-flash-image, gemini-2.0-flash \|
503	\| Text-to-Image \| wan2.7-image, wan2.7-image-pro, wan2.6-t2i, doubao-seedream-5.0/4.5/4.0, gpt-image-2 \|
504	\| Image-to-Image \| wan2.7-image, wan2.7-image-pro, doubao-seedream-5.0/4.5/4.0, gpt-image-2 \|
505	\| Video Generation \| wan2.7-i2v, wan2.6-i2v-flash, doubao-seedance-2.0 (Normal/Fast), kling-v3/v2.6/v2.5 \|
506
507	Model information is defined in `video-claw/video-claw/backend/models/config_model.py`. The frontend and Pipeline APIs filter models by capability tags, such as text generation, image generation, image-to-video, action transfer, and TTS.
508
509	</details>
510
511
512	# Artifacts and Storage
513
514	All tasks metadata and generated artifacts are stored under `video-claw/video-claw/backend/code/`.
515
516	<details>
517	<summary><b>Click to expand Storage Structure and Identifiers</b></summary>
518
519	## 📁 Storage Structure
520
521	```text
522	video-claw/video-claw/backend/code/
523	├── data/
524	│ ├── tasks/ # Pipeline task metadata (JSON)
525	│ └── sessions/ # AIGC-Claw session metadata (JSON)
526	└── result/
527	├── task/ # Pipeline generated artifacts (by Task ID)
528	│ └── <task_id>/ # e.g., 20260514_204946_961f95d9
529	│ ├── audio_xx.mp3 # Audio segments
530	│ ├── video_xx.mp4 # Video segments
531	│ ├── storyboard.json # Storyboard data
532	│ └── final.mp4 # Final combined video
533	├── image/ # AIGC-Claw generated images
534	│ └── <session_id>/ # By Session ID
535	│ ├── Assets/ # Character and setting assets
536	│ │ ├── characters/ # Character reference images
537	│ │ └── settings/ # Setting reference images
538	│ └── Scenes/ # Generated storyboard reference images
539	├── video/ # AIGC-Claw generated videos
540	│ └── <session_id>/ # By Session ID
541	└── script/ # AIGC-Claw generated script/storyboard data
542	```
543
544	## 🆔 Identifiers
545
546	- Task ID: Format: `YYYYMMDD_HHMMSS_RandomHash` (e.g., `20260514_204946_961f95d9`), used to uniquely identify a Pipeline task.
547	- Session ID: Millisecond-level timestamp (e.g., `1778810088325`), used to associate interaction context and generated images in the main workflow.
548
549	</details>
550
551	# 🙏 Acknowledgments
552
553	The idea and design of Video-Claw were inspired by [Pixelle-Video](https://github.com/AIDC-AI/Pixelle-Video), [AutoResearchClaw](https://github.com/aiming-lab/AutoResearchClaw), [huobao-drama](https://github.com/chatfire-AI/huobao-drama), [Flova](https://www.flova.ai), and [libtv-skills](https://github.com/libtv-labs/libtv-skills).
554
555
556	# 📚 Related Work
557
558	\| Framework \| Paper Information \|
559	\|:---:\|---\|
560	\| <img src="./FilmAgent-pics/framework.png" width="420" alt="FilmAgent framework"/> \| [SIGGRAPH Asia 2024] FilmAgent: Automating Virtual Film Production Through a Multi-Agent Collaborative Framework<br>Zhenran Xu, Jifang Wang, Longyue Wang, Zhouyi Li, Senbao Shi, Baotian Hu, Min Zhang<br>[[Paper](https://doi.org/10.1145/3681758.3698014)] [[GitHub](https://github.com/HITsz-TMG/Video-Claw/blob/main/FilmAgent.md)] \|
561	\| <img src="https://github.com/HITsz-TMG/Anim-Director/blob/main/Anim-Director/assets/visualeg.png" width="420" alt="Anim-Director result"/> \| [SIGGRAPH Asia 2024] Anim-Director: A Large Multimodal Model Powered Agent for Controllable Animation Video Generation<br>Yunxin Li, Haoyuan Shi, Baotian Hu, Longyue Wang, Jiashun Zhu, Jinyi Xu, Zhen Zhao, Min Zhang<br>[[Paper](https://doi.org/10.1145/3680528.3687688)] [[GitHub](https://github.com/HITsz-TMG/Anim-Director/tree/main/Anim-Director)] \|
562	\| <img src="https://raw.githubusercontent.com/HITsz-TMG/Anim-Director/main/AniMaker/assets/pipeline.png" width="420" alt="AniMaker pipeline"/> \| [SIGGRAPH Asia 2025] AniMaker: Multi-Agent Animated Storytelling with MCTS-Driven Clip Generation<br>Haoyuan Shi, Yunxin Li, Xinyu Chen, Longyue Wang, Baotian Hu, Min Zhang<br>[[Paper](https://doi.org/10.1145/3757377.3764009)] [[GitHub](https://github.com/HITsz-TMG/Anim-Director/tree/main/AniMaker)] \|
563
564
565	<p align="center">
566	<sub>Built with 🦞 by the Lychee Agent team</sub>
567	</p>
568

568 lines MARKDOWN