Reactor for ComfyUI — world models in your workflow

**Real-time video models in ComfyUI. Steer them live, or script every shot.** [🌐 Reactor](https://reactor.inc) · [📚 Docs](https://docs.reactor.inc) · [🛠️ Client SDKs](https://github.com/reactor-team/reactor-client-sdks) · [📖 Cookbook](https://github.com/reactor-team/reactor-cookbook)

--- A [ComfyUI](https://www.comfy.org) custom node pack for [Reactor](https://reactor.inc)'s real-time video models. Play a model live in a window you steer as it generates, or script a sequence of segments and render it to a video, all from a normal ComfyUI workflow.

The Reactor Realtime window driving a LingBot World 2 world with the keyboard
_{LingBot World 2 playing in the Reactor Realtime window, steered with W/A/S/D and the arrow keys.}

## Getting started **1. Install.** Clone this repo into `ComfyUI/custom_nodes/` and install its requirements into ComfyUI's Python: ```bash cd ComfyUI/custom_nodes git clone https://github.com/reactor-team/comfyui-reactor-api pip install -r comfyui-reactor-api/requirements.txt ``` **2. Add your API key.** Copy `config.ini.example` to `config.ini` and set your Reactor API key, or set `REACTOR_API_KEY` in ComfyUI's environment (it takes priority). **3. Copy the example inputs.** Copy `example_inputs/` into ComfyUI's `input` folder so the examples find their images. ## Examples Every example is in ComfyUI's template browser under **comfyui-reactor-api**, or drag a preview onto the canvas to load its workflow. Edit previews show the source on the left and the edit on the right. Each example runs in one of two ways: - **⚡ Real time** — the Reactor Realtime node opens a window where the model plays as it generates. Edit the prompt, drive the camera with the keyboard, feed it your webcam, or talk to an avatar. - **🎬 Render** — chain segment nodes into a sequence and Reactor Render turns it into a video. Other nodes can write the prompts, make the images, or branch the story. ### Explore worlds Walk through a world made from a single image, steering the camera live with the keyboard or scripting every move ahead of time.

_{Real-time world driven with the keyboard
LingBot World 2 · ⚡ Real time}

_{Scripted camera moves
LingBot World 2 · 🎬 Render}

### Edit video Restyle, re-dress or recast your camera or a video file from a single image, or edit it with a prompt. Source on the left, result on the right.

_{Real-time camera restyled from an image Vidu S2-Editing · ⚡ Real time}	_{Real-time camera dressed in an outfit from an image Vidu S2-Editing · ⚡ Real time}
_{Real-time camera recast as a character from an image Vidu S2-Editing · ⚡ Real time}	_{Real-time camera edited by a prompt X2 · ⚡ Real time}
_{Real-time video file edited by a prompt X2 · ⚡ Real time}	_{Restyle, re-dress and recast a video Vidu S2-Editing · 🎬 Render}

### Generate video Short clips with sound from a prompt with FastH3, or from pictures of a person and a place with H3 Reference Turbo Realtime. Stories across shots and cuts, one continuous scene steered by prompts, or a single shot that morphs, all from text.

_{Real-time clips with sound from a prompt FastH3 · ⚡ Real time}	_{Clips with sound, a shot then a cut FastH3 · 🎬 Render}
_{Real-time clips of a person and a place from two pictures H3 Reference Turbo Realtime · ⚡ Real time}	_{A person and a place from two pictures H3 Reference Turbo Realtime · 🎬 Render}

_{Real-time video from a prompt LongLive-2.0 · ⚡ Real time}	_{Nine segments of shots and cuts LongLive-2.0 · 🎬 Render}
_{One scene steered by prompts Helios · 🎬 Render}	_{A storm morphing around one shot Visko Orbis Stable · 🎬 Render}

### Avatars Talk with a character made from a photo: your voice goes in, video comes out. Or have a photo speak your script, one take at a time.

_{Real-time conversation, your voice in and video out Vidu S2-Avatar · ⚡ Real time}
_{Real-time takes of a photo speaking your script LTX · ⚡ Real time}	_{A photo speaks a script in two takes LTX · 🎬 Render}

## Models Read the model's prompt guide before writing prompts: it applies to the prompts you write here, and each example's note links it. ### Explore worlds | Model | What it does | Takes | Prompt guide | | --- | --- | --- | --- | | [LingBot World 2](https://docs.reactor.inc/model-api-reference/lingbot-world-2/overview) | LingBot with more camera control | first image, camera moves | [Prompt guide](https://docs.reactor.inc/model-api-reference/lingbot-world-2/prompt-guide) | | [LingBot](https://docs.reactor.inc/model-api-reference/lingbot/overview) | A world you walk through | first image, camera moves | [Prompt guide](https://docs.reactor.inc/model-api-reference/lingbot/prompt-guide) | ### Edit video | Model | What it does | Takes | Prompt guide | | --- | --- | --- | --- | | [Vidu S2-Editing](https://docs.reactor.inc/model-api-reference/vidu-s2-editing/overview) | Restyle, re-dress or recast from an image | video, image | [Prompt guide](https://docs.reactor.inc/model-api-reference/vidu-s2-editing/prompt-guide) | | [X2](https://docs.reactor.inc/model-api-reference/x2/overview) | Edit a video or camera from a prompt | video, optional image | [Prompt guide](https://docs.reactor.inc/model-api-reference/x2/prompt-guide) | | [Sana Streaming](https://docs.reactor.inc/model-api-reference/sana-streaming/overview) | Edit a video file from a prompt | video (Render only) | [Prompt guide](https://docs.reactor.inc/model-api-reference/sana-streaming/prompt-guide) | ### Generate video | Model | What it does | Takes | Prompt guide | | --- | --- | --- | --- | | [FastH3](https://docs.reactor.inc/model-api-reference/fast-h3/overview) | Short clips with sound, one per prompt | prompt, image on any segment | [Prompt guide](https://docs.reactor.inc/model-api-reference/fast-h3/prompt-guide) | | [H3 Reference Turbo Realtime](https://docs.reactor.inc/model-api-reference/h3-reference-to-video-turbo-realtime/overview) | Clips with sound, guided by up to nine pictures, which prompts call Picture 1, Picture 2, … in socket order | prompt, pictures | [Prompt guide](https://docs.reactor.inc/model-api-reference/h3-reference-to-video-turbo-realtime/prompt-guide) | | [Helios](https://docs.reactor.inc/model-api-reference/helios/overview) | One continuous scene steered by prompts | prompt, image on any segment | [Prompt guide](https://docs.reactor.inc/model-api-reference/helios/prompt-guide) | | [Visko Orbis Stable / Dynamic](https://docs.reactor.inc/model-api-reference/visko-orbis-stable/overview) | One unbroken shot that morphs, with sound | prompt, optional first image | [Stable](https://docs.reactor.inc/model-api-reference/visko-orbis-stable/prompt-guide), [Dynamic](https://docs.reactor.inc/model-api-reference/visko-orbis-dynamic/prompt-guide) | | [LongLive-2.0](https://docs.reactor.inc/model-api-reference/longlive-v2/overview) | A story across shots and cuts, from text | prompt | [Prompt guide](https://docs.reactor.inc/model-api-reference/longlive-v2/prompt-guide) | ### Avatars | Model | What it does | Takes | Prompt guide | | --- | --- | --- | --- | | [Vidu S2-Avatar](https://docs.reactor.inc/model-api-reference/vidu-s2-avatar/overview) | Talk with a character made from a photo | first image, persona | [Prompt guide](https://docs.reactor.inc/model-api-reference/vidu-s2-avatar/prompt-guide) | | [LTX](https://docs.reactor.inc/model-api-reference/ltx/overview) | A person from a photo speaks your script | first image, script | [Prompt guide](https://docs.reactor.inc/model-api-reference/ltx/prompt-guide) | **Not yet supported: HappyOyster.** ## Nodes These are the nodes this plugin adds to ComfyUI, all under **Reactor** in the node menu. | Node | What it does | | --- | --- | | **Reactor Realtime** | Runs a model in real time. Pick the model, queue it, and steer it in the window: edit the prompt and press Apply, drive the camera with W/A/S/D and the arrow keys, or talk to an avatar. Press Save to output the take as a video. | | **Reactor <model> Segment** | One segment of a scripted sequence, with that model's inputs. Chain segments through `sequence`. The first segment holds the model's settings. Since every input is a normal ComfyUI input, other nodes can write the prompts, make the images, or branch the story. | | **Reactor Render** | Renders a sequence and outputs a video. | | **Reactor Sequence Join** | Plays sequences for the same model one after another. | | **Reactor Sequence Visualizer** | Draws a sequence's segments and camera moves on a frame ruler. | | **Reactor Vidu S2-Avatar Reference** | Tags an image as an object, outfit or background for an avatar segment. | | **Reactor Camera Capture /
Microphone Capture** | Pick this browser's camera or microphone for Reactor Realtime. | ## Tips - **Getting the same take every time?** In Reactor Realtime, a fixed `seed` with unchanged inputs reuses the last take. Set it to randomize to record a new take each time. - **Looking for a take?** Reactor Realtime saves them to `reactor/realtime` in ComfyUI's `output` folder. Change it with the node's `filename_prefix`. - **Talking to an avatar?** Wear headphones, so the Vidu S2-Avatar character doesn't hear itself. - **Want to run more than one workflow at once?** ComfyUI runs one Reactor session at a time. Set `MAX_CONCURRENT` under `[Sessions]` in `config.ini` and restart to run more. ## Contributing Pull requests are welcome. Run the tests before opening one; they need no ComfyUI install and no Reactor API key: ```bash uv run pytest -q tests node --test "tests/web/*.test.mjs" ``` Sign off every commit (`git commit -s`): pull requests are checked for a [DCO](https://developercertificate.org/) sign-off. ## Licensing This repository is **Apache-2.0** licensed - see [`LICENSE`](LICENSE).

Reactor Sequence Join

Inputs (1)

Outputs (1)