ReShot
Copy the shot, not the actors. ReShot turns a reference video into a depth map, so Seedance or MiniMax H3 can repeat its choreography and camera moves — with your own characters in it.
复制走位，不复制演员。ReShot 把参考视频变成深度图，让 Seedance 或 MiniMax H3 照着它的动作和运镜再拍一遍——人物换成你的。
Open-sourced by Maosika 猫斯卡, the AI short-drama production system · 由 AI 短剧生产系统 猫斯卡 开源
top: the reference and its depth map · bottom: three takes generated from that depth map — two women, one rabbit; same moves, same camera
上排：参考片和它的深度图 · 下排：用这张深度图生成的三段片——两个女人、一只兔子；动作一样，镜头一样
Open the notebook, upload a clip, run, download depth.mp4 with the prompt line for your video model.
打开笔记本，上传视频，运行，下载 depth.mp4 和给视频模型的提示词。
pip install reshot
reshot # opens a page in your browser — drop the clip in, click, download
reshot reference.mp4 -o depth.mp4 --target seedance # or the command line
The web page (reshot with nothing after it): everything runs on your own computer; the prompt line for Seedance or MiniMax H3 is right there to copy.
网页版（只敲 reshot）：全部在你自己电脑上跑；给 Seedance / MiniMax H3 的提示词就在旁边，点一下复制。
The problem / 你遇到的问题
You have a clip whose fight, dance or camera move is exactly what you want in your own AI video. Feed it to the video model as a reference and it copies the faces, the clothes and the look along with the moves — and real people in the clip may not pass the content check at all. Describe the moves in words instead, and you get a different fight every time.
你看中了一段片子，那场打戏、那段舞、那个运镜正是你的 AI 视频想要的。直接当参考视频喂给模型，它会把脸、衣服、画风连同动作一起抄走，片子里是真人还可能过不了审；改用文字描述动作，模型每次给你的打法都不一样。
What ReShot does / ReShot 做什么
An .mp4 in, an .mp4 out. The output is a depth-map video: every frame grey, near is white, far is black. It keeps where everyone stands, how big they are, how they move and how the camera moves; it throws away faces, clothes, lighting and style. Give the grey video to your model as the reference and describe the people and the look in the prompt.
吃进一个 mp4，吐出一个 mp4。输出是深度图视频：每帧都是灰的，近白远黑。留下谁站在哪、谁大谁小、怎么动、镜头怎么走；扔掉脸、衣服、光线、画风。把灰片当参考视频交给模型，提示词里写人物和画风。
How to use it / 怎么用
| 1 | pip install reshot — or paste into Claude Code / Codex: "Install reshot from PyPI following https://raw.githubusercontent.com/maosika-ai/reshot/main/docs/AGENT_INSTALL.md and run a test clip". ComfyUI: install ComfyUI-ReShot.装；或者把这句粘给 Claude Code / Codex 让它装；ComfyUI 用户装节点包 |
| 2 | Type reshot — a page opens in your browser: drop the clip in, pick Seedance or MiniMax H3, click, download. Or the command line: reshot reference.mp4 -o depth.mp4 --target seedance — 24 fps, H.264, the Seedance reference-video spec out of the box.敲 reshot 打开网页：拖进去、选模型、点一下、下载。或者命令行 --target seedance 直接按 Seedance 参考视频规格出片 |
| 3 | Seedance 2.0 / 2.5: upload depth.mp4 as a reference video and prompt 参考@视频1的动作与运镜，顺序与视频保持一致。 + the people and the look.MiniMax H3: attach it as <Video 1>, a character sheet as <Picture 1>; define <Video 1> as "the fight choreography and camera movement shown in <Video 1>, a grey depth map…" and state that "its grey depth look is not transferred". The full prompts for the three takes are in docs/prompts.交给模型：Seedance 当参考视频上传并写「参考@视频1的动作与运镜」；MiniMax H3 挂成 <Video 1>、定妆图挂 <Picture 1>，明说灰色外观不要抄。三段片的完整提示词在 docs/prompts。 |
| 4 | Same six shots, same camera, your cast — the three takes above. Tips that made them work: keep the reference under 15 s; for MiniMax H3 shrink the depth map (--max-res 320) so the grey silhouette doesn't pull the face shape; keep the size ratio between characters as the depth map has it.出片。几条经验：参考片 15 秒以内；给 MiniMax H3 的深度图缩到 320（不然灰影会带脸型）；角色体量比例别跟深度图打架。 |
Works with the models you already use. / 你在用的模型都能接
--target | fps | size | for |
|---|---|---|---|
| seedance | 24 | ×16, ≥ 407,696 px, ≤ 15 s | Seedance 2.0 / 2.5 — attach as a reference video, prompt 参考@视频1的动作与运镜…. Meets the official reference-video spec (24 fps, H.264, size floor) out of the box. |
| h3 | 24 | ×32, ≤ 15 s | MiniMax H3 Fun ControlNet — depth condition |
| wan | 16 | ×16 | Wan 2.1 VACE (values from the docs, not yet verified end-to-end) |
| none | source | even | any depth ControlNet |
Steady by design. / 稳，是设计出来的
| Temporal consistency | a native video depth model — 32-frame windows, overlap, keyframe alignment — so walls don't breathe and floors don't flicker |
| Normalisation | once over the whole clip, never per frame |
| Frame rate | by timestamp: 30 → 24 really is 24 |
| Output size | cropped to the model's multiple — never padded; a black border reads to the generator as a far wall |
| Encoding | x264 CRF 12, no banding in the control signal |
| Memory | inference at model resolution: ~4 GB host RAM for 12 s of 720p, and it refuses up front instead of swapping |
Measured / 实测
| device | ms / frame | host RAM |
|---|---|---|
RTX 3080 Ti, --quality fast (default, model sees 644×364) | 34 · 3 GB VRAM | 2.3 GB |
RTX 3080 Ti, --quality full (model sees 924×518) | 83 · 11 GB VRAM | 3.5 GB |
RTX 4090, --quality full | 62 · 11 GB VRAM | 3.9 GB |
| Apple M2 Max, fp32 | ~500 | ~4 GB |
Yours to ship. / 拿去用
Apache-2.0 end to end — code and the default model (Video Depth Anything Small). CLI and a Python API with typed errors and exit codes; a fake backend runs the whole pipeline in milliseconds for your own tests. Runs on CUDA, Apple Silicon and CPU.
About Maosika 猫斯卡 / 关于猫斯卡
ReShot is open-sourced by Maosika 猫斯卡 (www.maosika.com), a professional AI video production system for short drama and short video. Maosika takes a story from a one-line idea to a finished vertical AI short drama: it writes the episodic script, designs the characters and scenes, generates consistent character sheets and scene images, and renders every shot with video models such as Seedance 2.0 / 2.5 and MiniMax H3 — a crew of digital specialists handling each step. Individual screenwriters, MCNs and short-drama companies use Maosika to produce AI short drama every day. ReShot is the depth-map step of that pipeline.
ReShot 由 猫斯卡（www.maosika.com）开源。猫斯卡是一套专业的 AI 视频自动生产系统，专注 AI 短剧、AI 短视频、AI 漫剧的全流程制作：从一句话创意出发，自动写出分集剧本，设计人物与场景，生成前后一致的角色定妆图和场景图，再用 Seedance 2.0 / 2.5、MiniMax H3 等视频模型把每一镜拍出来——每个环节都由对应的数字专家接手。个人编剧、MCN 机构和短剧公司每天都在用猫斯卡生产 AI 短剧。ReShot 就是这条生产线里的「深度图」环节。
The full story with the demo takes and prompts: the ReShot article on maosika.com · 中文文章