Extensions/Runware.ai ComfyUI Inference API Integration
ComfyUI Extension

Runware.ai ComfyUI Inference API Integration

Runware Inference API Integration for ComfyUI (No GPU Required).

By Runware·Created 2 years ago·Updated 26 days ago· 140
Runware/ComfyUI-Runware
Nodes380
On cloudLocal install
CategoryRunware/Audio/runware, Runware/Video/alibaba
Stars140
Updated26 days ago

Nodes (380)

ACE-Step v1.5 Base

Write a whole song from a prompt — ACE-Step makes lyrics, key, and BPM first-class inputs

Runware/Audio/runware
ACE-Step v1.5 Turbo

A music model that lets you set the BPM, key, and time signature

Runware/Audio/runware
ACE-Step v1.5 XL Base

The music node for people who actually want to steer the sampler

Runware/Audio/runware
ACE-Step v1.5 XL SFT

The full-quality ACE-Step, for the take that matters

Runware/Audio/runware
ACE-Step v1.5 XL Turbo

Full songs from a text prompt, fast

Runware/Audio/runware
HappyHorse-1.0

Alibaba's video editor, prompt your existing clips

Runware/Video/alibaba
HappyHorse 1.1

Alibaba's sleeper video model with the size presets

Runware/Video/alibaba
Qwen2.5-VL-3B-Instruct

Ask a vision model about your image, get a string back

Runware/Text/runware
Qwen2.5-VL-7B-Instruct

The image-to-text node that reads your images

Runware/Text/runware
Qwen3-TTS 1.7B Base

Clone a voice from a 3-second clip — Qwen3-TTS as a ComfyUI node

Runware/Audio/alibaba
Qwen3-TTS 1.7B CustomVoice

Give Your Video a Voice It Didn't Have

Runware/Audio/alibaba
Qwen3-TTS 1.7B VoiceDesign

Design a voice with a sentence — no reference clip required

Runware/Audio/alibaba
Qwen-Image

The open editing model, minus the GPU

Runware/Image/runware
Qwen-Image-2.0

The unified 7B, API-only for a reason

Runware/Image/alibaba
Qwen-Image-2.0-Pro

Qwen-Image 2.0 Pro — the API-only refresh of the open image flagship

Runware/Image/alibaba
Qwen-Image-2512

Alibaba's text-rendering flagship in the cloud

Runware/Image/alibaba
Qwen-Image-Edit

Describe the change, skip the mask

Runware/Image/runware
Qwen-Image-Edit-Plus

The better-behaved editor

Runware/Image/runware
Qwen-Image-Layered

Compositing layers out of thin air (PNG-keepers, this is for you)

Runware/Image/alibaba
Wan2.5-Preview

The Wan That Never Got Open Weights, Rendered

Runware/Video/runware
Wan2.5-Preview Image

Wan2.5-Preview Image — a Wan nobody can download, used through Runware

Runware/Image/runware
Wan2.6

Wan 2.6 without a 27B model on your desk

Runware/Video/alibaba
Wan2.6 Flash

Image-to-video with native audio, served from the cloud

Runware/Video/alibaba
Wan2.6 Image

Alibaba's image model with a surprising resolution ceiling

Runware/Image/alibaba
Wan2.7

Wan2.7 — the API-only Wan you couldn't run locally anyway

Runware/Video/alibaba
Wan2.7 Image

Wan 2.7 Image — the newest closed Wan, with a thinking-mode switch

Runware/Image/alibaba
Wan2.7 Image Pro

The Wan You Can Actually Run

Runware/Image/alibaba
Z-Image

Z-Image in the cloud — the 'SDXL 2.0' you can also afford to skip local for

Runware/Image/runware
Z-Image-Turbo

The 6B model that killed the FLUX 2 hype

Runware/Image/runware
Claude Fable 5

Claude Fable 5 — extended thinking with built-in prompt caching

Runware/Text/anthropic
Claude Haiku 4.5

Claude Haiku 4.5 in a node — the fast, cheap LLM for graph plumbing

Runware/Text/anthropic
Claude Opus 4.7

Claude Opus 4.7 — the reasoning tier for when the graph has to think

Runware/Text/anthropic
Claude Opus 4.8

Claude Opus 4.8 — the newest Opus, and the priciest node in the pack

Runware/Text/anthropic
Claude Sonnet 4.6

Claude Sonnet 4.6 — the middle child that usually wins

Runware/Text/anthropic
Exactly Illustrative (custom model)

Run an illustration checkpoint without owning a GPU

Runware/Custom models
FLUX.1 [dev] (custom model)

FLUX.1 [dev] and every FLUX.1 fine-tune, without the 24GB download

Runware/Custom models
FLUX.1 Kontext [dev] (custom model)

Kontext dev's edits, without the 12B-model hardware bill

Runware/Custom models
FLUX.1 [schnell] (custom model)

Any FLUX checkpoint, zero downloads

Runware/Custom models
Illustrious XL (custom model)

Danbooru tags, no VRAM

Runware/Custom models
NoobAI XL (custom model)

NoobAI XL without the vpred setup headache

Runware/Custom models
Pony Diffusion XL (custom model)

Every Pony checkpoint, one node

Runware/Custom models
Stable Diffusion 1.5 (custom model)

Any CivitAI SD 1.5 checkpoint, on Runware's GPUs

Runware/Custom models
Stable Diffusion XL (custom model)

Run any SDXL checkpoint without downloading a byte

Runware/Custom models
Stable Diffusion XL Lightning (custom model)

The fast SDXL rung, on any checkpoint

Runware/Custom models
Stable Diffusion XL Turbo (custom model)

SDXL Turbo in the cloud, with the whole SDXL ecosystem attached

Runware/Custom models
Z Image (custom model)

Z-Image — but any fine-tune you want, by AIR

Runware/Custom models
Z Image Turbo (custom model)

Z Image Turbo, but bring your own checkpoint

Runware/Custom models
ERNIE-Image

The text-in-image specialist that still lost

Runware/Image/baidu
ERNIE-Image-Turbo

The 8-step shortcut to text-in-image

Runware/Image/baidu
FLUX.1.1 [pro]

BFL's closed flagship, on your canvas

Runware/Image/bfl
FLUX.1.1 [pro] Ultra

Nano Banana, as a ComfyUI node

Runware/Image/bfl
FLUX.1 [dev]

FLUX.1 [dev] without the 24GB VRAM tax

Runware/Image/runware
FLUX.1 Expand [pro]

Growing your canvas, credibly

Runware/Image/bfl
FLUX.1 Fill [dev]

The Swiss-army editor

Runware/Image/runware
FLUX.1 Fill [pro]

Inpainting without the VAE churn

Runware/Image/bfl
FLUX.1 Kontext [dev]

The Edit-By-Instruction Model, In the Cloud

Runware/Image/runware
FLUX.1 Kontext [max]

The API-tier instruction editor

Runware/Image/bfl
FLUX.1 Kontext [pro]

Same Editing Magic, Fewer Knobs, Hosted

Runware/Image/bfl
FLUX.1 [schnell]

The fast, Apache-licensed FLUX

Runware/Image/runware
FLUX.2 [dev]

FLUX.2 [dev] without the 18–24GB VRAM bill

Runware/Image/runware
FLUX.2 [flex]

The Cheap, Fast Tier of the New Flux

Runware/Image/bfl
FLUX.2 [klein] 4B

The Apache-2.0 FLUX that runs on a lunch budget

Runware/Image/runware
FLUX.2 [klein] 4B Base

The size-distilled Flux 2, with actual dials

Runware/Image/runware
FLUX.2 [klein] 9B

The Editing Daily Driver, Without the 21GB

Runware/Image/runware
FLUX.2 [klein] 9B Base

The distilled Flux 2 that fixed the VRAM wall

Runware/Image/runware
FLUX.2 [klein] 9B KV

FLUX.2 Klein 9B — the consumer-grade Flux, served cloud-side

Runware/Image/runware
FLUX.2 [max]

The 32B flagship that doesn't need your VRAM

Runware/Image/bfl
FLUX.2 [pro]

The Flagship, Where Everything Is Decided For You

Runware/Image/bfl
FLUX Erase

Remove anything, and let the background heal

Runware/Image/bfl
FLUX Outpainting

Extend the Canvas, Keep the Subject

Runware/Image/bfl
FLUX Virtual Try-On

Person, garment, and a 4-step render

Runware/Image/bfl
BiRefNet Dis

The benchmark-bred edge specialist

Runware/Image/runware
BiRefNet General

Hair-level background removal, no VRAM spent

Runware/Image/runware
BiRefNet General 512×512 FP16

The cheap bulk cutout machine

Runware/Image/runware
BiRefNet HRSOD DHU

When the fine detail lives at high resolution

Runware/Image/runware
BiRefNet Massive TR DIS5K TR TES

The 'trained on everything' cutout model

Runware/Image/runware
BiRefNet Matting

The hair-saving cutout, minus the model download

Runware/Image/runware
BiRefNet Portrait

Background removal that survives hair

Runware/Image/runware
BiRefNet v1 Base

The background remover people actually reach for

Runware/Image/runware
BiRefNet v1 Base – COD

The camouflage hunter you'll rarely need

Runware/Image/runware
Bria 3.2

The commercially-safe generator with the moderation trap

Runware/Image/bria
Bria FIBO

Bria FIBO without the 16GB download — cloud text-to-image that's built for editing

Runware/Image/bria
Bria FIBO Edit

Iterative image editing the way FIBO was meant to be used

Runware/Image/bria
Bria Fibo Edit Tools

Seven image editors in one node, and one of them is a time machine

Runware/Image/bria
Bria FIBO Lite

Bria FIBO Lite — the version of FIBO that took away your dials

Runware/Image/bria
Bria Image Increase Resolution

The one-knob cloud upscaler — 2x or 4x, zero VRAM

Runware/Image/bria
Bria Image Replace Background

Product-photo background swaps without cutting out anything by hand

Runware/Image/bria
Bria RMBG v2.0

The zero-settings background remover — Bria RMBG v2.0, one input, one output

Runware/Image/bria
Runware Bria Rmbg V2 0 Open

RMBG v2.0 Open — the same cutout, under Runware's own catalog

Runware/Image/runware
Bria Video Background Removal

Video background removal, and the one thing that will trip you first

Runware/Image/bria
Bria Video Eraser

Object erasure in video — make the thing disappear, keep the audio

Runware/Video/bria
Bria Video Increase Resolution

2x or 4x video upscaling on the cloud — with a frames-not-video twist

Runware/Image/bria
Runware Accelerator Options

The speed levers — TeaCache, DeepCache, and friends for cloud inference

Runware/Params
Runware Audio Settings

Audio Settings — the builder with three knobs and a lot of defaults

Runware/Params
Runware ControlNet

Cloud ControlNet — spatial structure, minus the local model files

Runware/Params
Runware Embeddings

Embeddings — textual inversion for Runware models, if you've got one

Runware/Params
Runware Import Model

Import Model — how your own checkpoint ends up on Runware's GPUs

Runware/Params
Runware IP-Adapter

The IP-Adapter you don't have to install

Runware/Params
Runware LoRA

The builder node that makes 'lora' sockets usable

Runware/Params
Runware Messages

Messages — build a conversation, one node at a time

Runware/Params
Runware Outpaint

Outpaint — extend the canvas, let the model invent what's beyond the edge

Runware/Params
Runware PhotoMaker

PhotoMaker — keep the face, change everything else

Runware/Params
Runware PuLID

PuLID in the cloud — face identity without the InsightFace install

Runware/Params
Runware Reference Images

Reference Images — the @-tag system that powers try-on and lip-sync

Runware/Params
Runware Reference Videos

Feed a Clip Into a Video Model

Runware/Params
Runware Reference Voices

Clone a Voice From a Clip — and Please, Get the Transcript Right

Runware/Params
Runware Refiner

Handing the Last Steps to a Refiner

Runware/Params
Runware Speech

The Builder That Feeds Every TTS Model

Runware/Params
Runware Ultralytics

YOLO Finds the Faces, Diffusion Fixes Them

Runware/Params
OmniHuman-1

One Photo, One Audio Clip, One Animated Human

Runware/Video/bytedance
OmniHuman-1.5

Same Party Trick, Now With a Mask and a Fast Lane

Runware/Video/bytedance
Seedance 1.0 Pro

Text-to-Video That Respects Your Camera

Runware/Video/bytedance
Seedance 1.0 Pro Fast

The Iteration Mode of the Same Video Model

Runware/Video/bytedance
Seedance 1.5 Pro

Draft Cheap, Promote the Winner, Ship With Sound

Runware/Video/bytedance
Seedance 2.0

The full-quality ByteDance video model, in the graph

Runware/Video/bytedance
Seedance 2.0 Fast

The speed tier of ByteDance's video model

Runware/Video/bytedance
Seedance 2.0 Mini

ByteDance's video model as a ComfyUI node

Runware/Video/bytedance
Seed Audio 1.0

One Prompt for Voices, SFX, Music, and Dialogue

Runware/Audio/bytedance
Seedream 4.0

ByteDance's Flagship Image Model, Without the Weights Problem

Runware/Image/bytedance
Seedream 4.5

ByteDance's flagship, minus the API plumbing

Runware/Image/bytedance
Seedream 5.0 Lite

ByteDance's image model, fast tier, sequence mode

Runware/Image/bytedance
Seedream 5.0 Pro

ByteDance's Seedream 5.0 Pro — great text rendering, huge canvas, zero local VRAM

Runware/Image/bytedance
BytePlus Video Enhancement Pro

Crank a video to 4K in the cloud, BytePlus edition

Runware/Image/bytedance
BytePlus Video Enhancement Standard

BytePlus video enhancement with a style dial for the source

Runware/Image/bytedance
ByteDance Video Upscaler

The Video Upscaler That Returns an Image, and Other Schema Quirks

Runware/Image/bytedance
CCSR

CCSR — the detail-restoring upscaler, without the 12GB+ VRAM tax

Runware/Image/runware
Clarity

Runware's Own Upscaler, Prompt Included

Runware/Image/runware
Content Safety

A pre-flight NSFW check that costs pennies

Runware/Text/runware
ControlNet Preprocess Canny

The edge-map factory for cloud ControlNet

Runware/Image/runware
ControlNet Preprocess Depth

The Preprocessor That Sits Before Your ControlNet

Runware/Image/runware
ControlNet Preprocess Lineart

The Character-Consistency Shortcut

Runware/Image/runware
ControlNet Preprocess Lineart Anime

Clean Lines for Cleaner Characters

Runware/Image/runware
ControlNet Preprocess MLSD

Straight-line ControlNet conditions, without the annotator stack

Runware/Image/runware
ControlNet Preprocess NormalBae

ControlNet for People Who Care About Light

Runware/Image/runware
ControlNet Preprocess OpenPose

Make Your Character Do the Reference's Move

Runware/Image/runware
ControlNet Preprocess Scribble

Turn a doodle into a ControlNet condition

Runware/Image/runware
ControlNet Preprocess Seg

Control layout with a segmentation map, not a sketch

Runware/Image/runware
ControlNet Preprocess Shuffle

Steal an Image's Ingredients, Not Its Composition

Runware/Image/runware
ControlNet Preprocess SoftEdge

The Forgiving ControlNet Preprocessor

Runware/Image/runware
ControlNet Preprocess Tile

The tile preprocessor that powers cloud upscale workflows

Runware/Image/runware
Aurora v1

Feed it a still and a soundtrack, get a clip

Runware/Video/creatify
Aurora v1 Fast

The same audio-to-video, on a budget clock

Runware/Video/creatify
Runware (custom)

The escape hatch for models the pack hasn't met yet

Runware/Custom
DeepSeek-V4-Flash

An LLM inside your ComfyUI graph

Runware/Text/deepseek
DeepSeek-V4-Pro

An LLM Living Inside Your ComfyUI Graph

Runware/Text/deepseek
Dia2 2B

The bigger, fiddlier TTS model for when you want control

Runware/Audio/runware
Exactly Illustrative Training

A LoRA-style tune-up that rents the GPU

Runware/Training/exactly
Exactly Bright Pulse

Reference-driven photos, zero prompt required

Runware/Image/exactly
Exactly Distant Reality

Reference in, out-of-focus world out

Runware/Image/exactly
Exactly Extreme Contrast

For the shots that snap

Runware/Image/exactly
Exactly Grain Film Look

The Film-Grain Node That Doesn't Even Have a Prompt Box

Runware/Image/exactly
Exactly Journey

The all-rounder of the reference-photo family

Runware/Image/exactly
Exactly Warm Light

Golden-hour vibes, from a node with no prompt

Runware/Image/exactly
Fish Audio S2.1 Pro

Give your ComfyUI workflow a voice — cloud TTS that can clone one

Runware/Audio/fishaudio
FLUX.1 [dev] SRPO

The dev-tuned FLUX that ditched the plastic look

Runware/Image/runware
Runware Get

Pull one field out of a JSON firehose

Runware/Params
Gemini 3.1 Flash Lite

The cheapest LLM in the graph

Runware/Text/google
Gemini 3.1 Flash TTS

Google-grade speech, straight into the graph

Runware/Audio/google
Gemini 3.1 Pro

The thinking model for the hard parts of your workflow

Runware/Text/google
Gemini 3.5 Flash

The one that searches the web and answers in JSON

Runware/Text/google
Gemini 3 Flash

An LLM in your graph, no API account of your own

Runware/Text/google
Gemini Omni Flash

Google's Video Model That Takes In Everything

Runware/Video/google
Gemma 4 31B

A 31B reasoning model inside your ComfyUI graph — with no GPU of your own

Runware/Text/google
Nano Banana

Google's image gen, minus the local GPU

Runware/Image/google
Nano Banana 2

The speed-quality hybrid that finally takes video refs

Runware/Image/google
Nano Banana 2 Lite

Nano Banana 2 Lite is Google's image model inside your ComfyUI graph

Runware/Image/google
Nano Banana Pro

Google's Gemini image model, no Google account needed

Runware/Image/google
Veo 3.1

Veo 3.1 — the frontier video model — from a ComfyUI node

Runware/Video/google
Veo 3.1 Fast

Veo 3.1 Fast drops Google's best video into a ComfyUI node — audio included

Runware/Video/google
Veo 3.1 Lite

Google's Best Video Model, on a Budget

Runware/Video/google
HeyGen Avatar IV

HeyGen Avatar IV turns a still photo into a talking presenter

Runware/Video/heygen
HeyGen Avatar V

A presenter, on demand, straight from a dropdown

Runware/Video/heygen
HeyGen Video Agent

A talking-head avatar in ComfyUI — 1,288 avatars on tap

Runware/Video/heygen
HiDream-I1 Dev

HiDream-I1 Dev — the MIT-licensed prompt-follower, minus the 17B weight problem

Runware/Image/runware
HiDream-I1 Fast

The prompt-adherence specialist, at iteration speed

Runware/Image/runware
HiDream-I1 Full

HiDream-I1 at full precision, no 40GB download

Runware/Image/runware
Rodin Gen-2

Rodin Gen-2 puts a 3D modeler in your ComfyUI graph

Runware/3D/hyper3d
Ideogram 2.0

The text-on-image specialist, parked inside your graph

Runware/Image/ideogram
Ideogram 2.0 Remix

Ideogram 2.0 Remix edits your image with a strength dial

Runware/Image/ideogram
Ideogram 2a

Ideogram 2a is the lean cloud option for text-on-image

Runware/Image/ideogram
Ideogram 2a Remix

The lightweight remix node that keeps your text

Runware/Image/ideogram
Ideogram 3.0

Character consistency plus 63 style presets in one node

Runware/Image/ideogram
Ideogram 3.0 Edit

Ideogram 3.0 Edit

Runware/Image/ideogram
Ideogram 3.0 Reframe

Ideogram 3.0 Reframe

Runware/Image/ideogram
Ideogram 3.0 Remix

Ideogram 3.0 Remix

Runware/Image/ideogram
Ideogram 3.0 Replace Background

Product shots without the photo studio

Runware/Image/ideogram
Ideogram 4.0

The text-in-image champ, with its own prompt schema

Runware/Image/ideogram
Ideogram 4.0q

The text-in-image king, without the 24GB download

Runware/Image/ideogram
Ideogram 4.0 Remix

The hosted Ideogram 4 remix — no weights, no filter fights

Runware/Image/ideogram
Ideogram Layerize Text

Pull text out of an image like it's a layer

Runware/Image/ideogram
Ideogram Object Remover

Mask it, delete it, keep the rest pixel-identical

Runware/Image/ideogram
ImagineArt 1.5

The cheap-and-cheerful text-to-image node

Runware/Image/imagineart
ImagineArt 1.5 Pro

ImagineArt 1.5 Pro

Runware/Image/imagineart
ImagineArt 2.0

ImagineArt 2.0

Runware/Image/imagineart
Inworld TTS-1.5 Max

Voice that doesn't sound like it's reading a spec sheet

Runware/Audio/inworld
Inworld TTS-1.5 Mini

The fast, cheap voice for first passes

Runware/Audio/inworld
Inworld Realtime TTS-2

Realtime voice, minus the sliders

Runware/Audio/inworld
Kandinsky 5.0 Image Lite

The Runware node that actually feels like local ComfyUI

Runware/Image/runware
KlingAI 1.0 Pro

The cloud video model that hides its GPU from you

Runware/Video/klingai
KlingAI 1.0 Standard

The value tier for testing cloud video cheaply

Runware/Video/klingai
KlingAI 1.5 Pro

Image-to-video that finally feels like a production tool

Runware/Video/klingai
KlingAI 1.5 Standard

Cheaper I2V when the clip isn't the hero

Runware/Video/klingai
KlingAI 1.6 Pro

The generation people actually talked about

Runware/Video/klingai
KlingAI 1.6 Standard

The 1.6 look on a value-tier budget

Runware/Video/klingai
KlingAI 2.0 Master

The ceiling the open-source world was chasing

Runware/Video/klingai
KlingAI 2.1 Master

The flagship that fixed the details

Runware/Video/klingai
KlingAI 2.1 Pro

Flagship image-to-video with the frame wired in

Runware/Video/klingai
KlingAI 2.1 Standard

The flagship's cheaper image-to-video half

Runware/Video/klingai
KlingAI 2.5 Turbo Pro

Speed as a feature, motion as the product

Runware/Video/klingai
KlingAI 2.5 Turbo Standard

Kling 2.5 Turbo on your ComfyUI canvas, without the 80GB checkpoint

Runware/Video/klingai
KlingAI Avatar 2.0 Pro

A talking head from one photo and an audio file

Runware/Video/klingai
KlingAI Avatar 2.0 Standard

Talking-head video at a price you can loop

Runware/Video/klingai
Kling IMAGE 3.0

Kuaishou's image model, on your ComfyUI canvas

Runware/Image/klingai
Kling IMAGE O1

The reference-driven image model with 'from input' sizes

Runware/Image/klingai
Kling IMAGE O3

The one that wants your exact pixel dimensions

Runware/Image/klingai
KlingAI Lip-Sync

New audio for an existing video, done in the cloud

Runware/Video/klingai
Kling VIDEO 2.6 Pro

The everything-input video node with sound

Runware/Video/klingai
Kling VIDEO 2.6 Standard

Reference-driven video, trimmed for cost

Runware/Video/klingai
Kling VIDEO 3.0 4K

15-second clips at real resolution, with sound

Runware/Video/klingai
Kling VIDEO 3.0 Omni 4K

Reference-video editing at cinema resolution, no GPU required

Runware/Video/klingai
Kling VIDEO 3.0 Omni Pro

The middle tier that most reference-video edits deserve

Runware/Video/klingai
Kling VIDEO 3.0 Omni Standard

The cheap way to iterate on reference-video edits

Runware/Video/klingai
Kling VIDEO 3.0 Pro

The text-to-video workhorse, with a negative prompt and real controls

Runware/Video/klingai
Kling VIDEO 3.0 Standard

The budget text-to-video node for prompt roulette

Runware/Video/klingai
Kling VIDEO 3.0 Turbo

Kling 3.0 Turbo's multi-shot prompt is the hidden superpower most people miss

Runware/Video/klingai
Kling VIDEO O1 Pro

The 'thinking' tier for edits that need more than a prompt pass

Runware/Video/klingai
Kling VIDEO O1 Standard

The cheap lane for learning what O1 edits can do

Runware/Video/klingai
Krea 2 Large

The hype model, served hot via API with no 24GB VAE hunt

Runware/Image/krea
Krea 2 Medium

The cheaper tier for testing prompts against the hype model

Runware/Image/krea
Krea 2 Medium Turbo

The hosted Krea 2 that follows prompts better than the open weights

Runware/Image/krea
Krea 2 Raw

The undistilled base, served over an API with real sampler dials

Runware/Image/krea
Krea 2 Turbo

The 12B base everyone's building on

Runware/Image/krea
FLUX.1 Krea [dev]

The Aesthetic Flux, With Every Socket Attached

Runware/Image/runware
LTX-2

The synchronized-audio video model, now a single ComfyUI node

Runware/Video/lightricks
LTX-2.3

The release that made LTX competitive, now without the local VRAM tax

Runware/Video/lightricks
LTX-2.3 Fast

The distilled tier that trades steps for speed, without losing the LoRA socket

Runware/Video/lightricks
LTX-2 Fast

The presets-only tier that makes one question go away

Runware/Video/lightricks
LTX-2 Pro

The audio-synced video model that Lightricks bet the family on

Runware/Video/lightricks
LTX-2 Retake

Fix one bad segment of a video, not the whole clip — LTX-2 Retake, the cloud way

Runware/Video/lightricks
Llama 3.1 8B Prompt Enhancer

A cheap way to stop writing bad prompts yourself

Runware/Text/runware
Runware Load Image (URL)

The tiny node that feeds every other Runware node

Runware/Params
Ray3.2

The reframe node for people who think in canvases, not clips

Runware/Video/luma
UNI-1

The image model that will Google for you before it draws

Runware/Image/luma
UNI-1 Max

The bigger-budget sibling for when the first pass isn't enough

Runware/Image/luma
MediaPipe Eyes + Lips Mesh

A cloud mask maker for the face-fix pipeline

Runware/Image/runware
MediaPipe Eyes Mesh

Want a clean mask of just the eyes? This cloud MediaPipe node hands you one

Runware/Image/runware
MediaPipe Face Full

The License-Clean Face Mask You've Been Missing

Runware/Image/runware
MediaPipe Face Mesh

Full-face masks in one click, from the cloud

Runware/Image/runware
MediaPipe Face Mesh Eyes Only

An eyes-only face mask, via the API — the boring node that earns its keep

Runware/Image/runware
MediaPipe Face Short

The tighter face mask, for edits that shouldn't wander

Runware/Image/runware
MediaPipe Lips Mesh

A lips-only mask, without drawing a single pixel

Runware/Image/runware
MediaPipe Nose + Eyes Mesh

One mask for the eyes and nose — the 'what are you looking at' region

Runware/Image/runware
MediaPipe Nose + Lips Mesh

Nose + lips in one mask, for the middle of the face

Runware/Image/runware
MediaPipe Nose Mesh

A nose-only mask — yes, this is a real use case

Runware/Image/runware
Memories Video Age Detection

Ask a model how old the person in your video is

Runware/Text/memories
Memories Video Captioning

Auto-caption a video in one node — no local vision model

Runware/Text/memories
Meshy-6

From a photo to a textured GLB, without touching Blender

Runware/3D/meshy
LLaVA-1.6-Mistral-7B

A vision-language model that talks about your image, not just describes it

Runware/Text/runware
SAM 3D Objects

The 'turn this one object into a 3D model' node

Runware/3D/meta
TRELLIS.2

Image to a real, textured 3D mesh — without the 3D stack

Runware/3D/microsoft
MiniMax 01

MiniMax 01 — the cloud video model with a prompt optimizer on tap

Runware/Video/minimax
MiniMax 01 Director

When your prompt is a shot list

Runware/Video/minimax
MiniMax 01 Live

The Video Node With No Required Inputs At All

Runware/Video/minimax
MiniMax Hailuo 02

The earlier MiniMax video tier that still earns its spot

Runware/Video/minimax
MiniMax Hailuo 2.3

Cloud video gen that isn't a local-VRAM nightmare

Runware/Video/minimax
MiniMax Hailuo 2.3 Fast

The image-to-video express lane

Runware/Video/minimax
MiniMax M2.7

Yes, this pack does text LLMs too

Runware/Text/minimax
MiniMax M2.7 Highspeed

A long-context LLM in the graph

Runware/Text/minimax
MiniMax M3

MiniMax M3 without a GPU

Runware/Text/minimax
MiniMax Music 2.6

Full songs from a prompt in ComfyUI, no GPU, no model files

Runware/Audio/minimax
MiniMax Music Cover

Restyle an existing song into a cover version, in-graph

Runware/Audio/minimax
MiniMax Speech 2.8

Character voices and dialogue without a TTS model download

Runware/Audio/minimax
Mirelo SFX 1.5

Type 'door creak' and get actual door-creak audio

Runware/Audio/mirelo
Mirelo SFX 1.6

Sound effects from a prompt, because you shouldn't have to record a door

Runware/Audio/mirelo
Kimi K2.6

Kimi K2.6 — a 1M-token-reasoning model as a ComfyUI node

Runware/Text/moonshotai
Object Eraser

The clean way to delete things from a photo

Runware/Image/runware
Open Age Detection

A one-input node that guesses age from a face

Runware/Text/runware
OpenAI CLIP ViT-L/14

A caption for any image, from the text encoder that ran SD 1.5

Runware/Text/runware
GPT-5.4

A full-featured GPT-5.4 chat node, straight into ComfyUI

Runware/Text/openai
GPT-5.4 Mini

The everyday GPT-5.4 — fast, cheap, same controls

Runware/Text/openai
GPT-5.4 Nano

When the job is tiny and you're calling it a lot

Runware/Text/openai
GPT-5.4 Pro

The heavy thinker with the training wheels removed

Runware/Text/openai
GPT-5.5

The newest model in the pack, JSON-native and thinking by default

Runware/Text/openai
GPT-5 Mini

The compact model that still speaks JSON

Runware/Text/openai
GPT-5 Nano

The Runware text node

Runware/Text/openai
GPT Image 1

The text-rendering king, without the OpenAI subscription

Runware/Image/openai
GPT Image 1.5

Same socket, sharper output

Runware/Image/openai
GPT Image 1 Mini

The cheap seat on the text-rendering train

Runware/Image/openai
GPT Image 2

GPT Image 2 in your ComfyUI graph — with moderation dialed to 'low'

Runware/Image/openai
Sora 2

OpenAI's video model without the API plumbing

Runware/Video/openai
Sora 2 Pro

The 'turn the dial up' Sora node

Runware/Video/openai
Picsart Image Vectorizer

Turn raster art into a real SVG

Runware/Vector/picsart
PixVerse LipSync

Make any video talk, and mean it

Runware/Video/pixverse
PixVerse Modify

Edit an existing video, not just make one

Runware/Video/pixverse
PixVerse V3.5

The workhorse text-to-video, effects included

Runware/Video/pixverse
PixVerse V4

The camera gets a director

Runware/Video/pixverse
PixVerse V4.5

Where the TikTok effect templates live

Runware/Video/pixverse
PixVerse V5

The quality pick of the family

Runware/Video/pixverse
PixVerse V5.5

The version where PixVerse got audio and 'thinking'

Runware/Video/pixverse
PixVerse V5.6

The thinking, sound-on refresh

Runware/Video/pixverse
PixVerse V5 Fast

Iterate like you mean it

Runware/Video/pixverse
PixVerse V6

Hosted video gen in your graph

Runware/Video/pixverse
P-Image

Pruna's take on text-to-image, minus the GPU

Runware/Image/prunaai
P-Image-Edit

The image editor that keeps the prompt simple

Runware/Image/prunaai
P-Image-Try-On

A fitting room in a node

Runware/Image/prunaai
P-Image Upscale

Outsource the upscale, keep the quality

Runware/Image/prunaai
P-Video

Text-to-video with the sound already baked in

Runware/Video/prunaai
P-Video-Animate

Bring a reference to life

Runware/Video/prunaai
P-Video-Avatar

Make a still image talk — P-Video-Avatar turns one face into a lip-synced clip

Runware/Video/prunaai
P-Video-Replace

Swap the subject, keep the performance

Runware/Video/prunaai
Qwen2.5-VL-7B Age Detector

How old is this person, for real? Runware's Qwen2.5-VL age check

Runware/Text/runware
Real-ESRGAN

The old reliable upscaler, now with no VRAM at all

Runware/Image/runware
Recraft V4

The designer's model, now a plain ComfyUI node

Runware/Image/recraft
Recraft V4.1

The refresh, and when it's actually worth it

Runware/Image/recraft
Recraft V4.1 Pro

The better rendering, at a per-image price

Runware/Image/recraft
Recraft V4.1 Utility

The boring Recraft that does real work

Runware/Image/recraft
Recraft V4.1 Utility Pro

The pristine product shot, on demand

Runware/Image/recraft
Recraft V4 Pro

The flagship design model, one node away

Runware/Image/recraft
Recraft V4 Pro Vector

Prompt in, honest-to-goodness SVG out

Runware/Vector/recraft
Recraft V4 Vector

The prompt-to-SVG node that just works

Runware/Vector/recraft
Recraft Vectorize

Turn the raster you already have into an SVG

Runware/Vector/recraft
RemBG v1.4

Cutouts without the model downloads

Runware/Image/runware
Reve 2.1

A closed 4K photorealism model you can reach from a ComfyUI node

Runware/Image/reve
Juggernaut Base Flux

The realism brand, moved to Flux and the cloud

Runware/Image/rundiffusion
Juggernaut Lightning Flux

Near-instant realism, no GPU

Runware/Image/rundiffusion
Juggernaut Pro Flux

Juggernaut Pro Flux, no 24GB VRAM required

Runware/Image/rundiffusion
Juggernaut Z

The photorealism default, now on Z-Image

Runware/Image/rundiffusion
FLUX.1 [dev] Style LoRA Training

Train a FLUX.1 style LoRA on someone else's GPU

Runware/Training/runware
FLUX.2 [klein] 4B Style LoRA Training

Train a FLUX.2 Klein 4B style LoRA without touching your GPU

Runware/Training/runware
FLUX.2 [klein] 9B Style LoRA Training

FLUX.2 Klein 9B style LoRA training, with the VRAM math handled for you

Runware/Training/runware
Qwen-Image Style LoRA Training

Train a Qwen-Image style LoRA in the cloud — your GPU gets the day off

Runware/Training/runware
Z-Image Style LoRA Training

Train a Z-Image style LoRA from ComfyUI — no GPU, no Kohya

Runware/Training/runware
Runway Aleph

Video-to-video from the company that made SD 1.5

Runware/Video/runway
Runway Aleph 2.0

Runway Aleph 2.0 video edits, from inside your graph

Runware/Video/runway
Runway Gen-4.5

The image-to-video node that only needs frames

Runware/Video/runway
Runway Gen-4 Image

Runway Gen-4 images, without the Runway website

Runware/Image/runway
Runway Gen-4 Image Turbo

Reference-driven images on the fast lane

Runware/Image/runway
Runway Gen-4 Turbo

Frames in, motion out, in ComfyUI

Runware/Video/runway
SkyReels V4

SkyReels V4 — the video model that can add its own sound

Runware/Video/skywork
Riverflow 2.0 Fast

The design model the internet is sleeping on

Runware/Image/sourceful
Riverflow 2.0 Pro

Riverflow 2.0 Pro — the cloud image node with built-in upscaling reference

Runware/Image/sourceful
Riverflow 2.5 Fast

Thinking, transparency, and a scoring rubric

Runware/Image/sourceful
Riverflow 2.5 Pro

The 4K tier with an extra level of thinking

Runware/Image/sourceful
Stable Diffusion 3

A real Stable Diffusion 3 node that runs on someone else's GPU

Runware/Image/runware
Stable Diffusion XL v1.0 VAE Fix

The boring checkpoint that saves you a headache

Runware/Image/civitai
Stable Diffusion Latent Upscaler

The classic latent upscaler, now with nothing to download

Runware/Image/runware
SwinIR

The Cloud Upscaler That Skips the Whole Local Dance

Runware/Image/runware
Runware Sync 3

Make a video's mouth match any audio, without running a lip-sync model locally

Runware/Video/sync
lipsync-2

A lipsync node that actually takes your audio

Runware/Video/sync
lipsync-2-pro

Make the mouth match the audio, without a local model

Runware/Video/sync
react-1

Re-animate a face to a new performance — the deepfake-adjacent node

Runware/Video/sync
Hunyuan 3D 3.1 Pro

Generating .glb files from your ComfyUI graph

Runware/3D/tencent
Hunyuan 3D 3.1 Rapid

Prompt-to-GLB without owning a 24GB GPU

Runware/3D/tencent
Topaz Labs Starlight Precise 2.5

Topaz's generative video upscaler, as a cloud ComfyUI node

Runware/Image/topazlabs
Tripo 3D v3.1

A GLB out of a text prompt (or an image)

Runware/3D/tripo
TwinFlow Z-Image-Turbo

The local-realism favorite, hosted and fast

Runware/Image/runware
Runware Upload Image

The boring node that makes the whole cloud pack work

Runware/Params
VEED Fabric 1.0

Turn a photo and an audio clip into a talking presenter, in the cloud

Runware/Video/veed
Vidu 1.5

Vidu 1.5 — the video node with a BGM switch and movement dial

Runware/Video/vidu
Vidu 2.0

The pragmatic text-to-video pick on Runware

Runware/Video/vidu
Vidu Q1 (Video)

The style-preset video model that hides its prompt box behind the fun part

Runware/Video/vidu
Vidu Q1 Classic

The image-to-video node with no prompt box — and that's the point

Runware/Video/vidu
Vidu Q1 (Image)

The odd one out — a Vidu image generator that needs a reference photo

Runware/Image/vidu
Vidu Q2 Pro

The generation-two Vidu that finally lets you dial in fractional seconds

Runware/Video/vidu
Vidu Q2 Turbo

The fast sibling of the Vidu Q2 line, prompt required

Runware/Video/vidu
Vidu Q3

The Chinese studio's video model, no account required

Runware/Video/vidu
Vidu Q3 Turbo

The newest Vidu, with native audio and a demand for explicit dimensions

Runware/Video/vidu
ViT Age Classifier

A one-socket 'how old is this person' node

Runware/Text/runware
Grok 4.3

A full LLM node that needs a messages builder to talk to

Runware/Text/xai
Grok Imagine Image

XAI's generator as a plain ComfyUI node

Runware/Image/xai
Grok Imagine Image Quality

The least-censored image model you can wire straight into ComfyUI

Runware/Image/xai
Grok Imagine Video

XAI's Video Model, Now a ComfyUI Node

Runware/Video/xai
Grok Imagine Video 1.5

Turn a frame into motion with xAI's Grok Imagine Video

Runware/Video/xai
xAI Text-to-Speech

The pack's shared speech engine, exposed as a bare audio node

Runware/Audio/xai
YOLOv8n Face

The tiny detector that feeds your face workflows

Runware/Image/runware
YOLOv8n Hand

Cloud hand detection that hands you a mask, not a bounding box

Runware/Image/runware
YOLOv8n Person Seg

Person silhouettes as masks, no local segmentation stack

Runware/Image/runware
YOLOv8s Face

A face mask in seconds, for fix-the-face workflows without the local model

Runware/Image/runware
YOLOv8s Person Seg

The small-model person segmenter for when nano isn't cutting it

Runware/Image/runware
GLM-5.1

Z.ai's newest model, tool-calling enabled, with a bigger token budget

Runware/Text/zai
Readme

ComfyUI-Runware

Every Runware model as a ComfyUI node: image, video, audio, 3D, and text, all running in the cloud. No local GPU and no per-model setup. The whole catalog shows up in your node menu, and each node's widgets are the model's real parameters.

Full guide: https://runware.ai/docs/platform/comfyui

Install

ComfyUI Manager (recommended): open the Custom Nodes Manager, search Runware, install, and restart.

Manual:

cd ComfyUI/custom_nodes
git clone https://github.com/Runware/ComfyUI-Runware
pip install -r ComfyUI-Runware/requirements.txt

Restart ComfyUI.

API key

Create a key in the dashboard, then provide it one of these ways:

  • ComfyUI Settings → Runware API key: paste it in the UI, no terminal needed.
  • RUNWARE_API_KEY environment variable: overrides the Settings field, useful for servers.
  • Runware CLI: run runware auth login once and the nodes reuse the stored key.

Quick start

  1. Double-click the canvas and search a model by name (e.g. FLUX.2 [dev]), or browse Runware/Image.
  2. Type your positivePrompt and set the dimensions.
  3. Wire the node's IMAGE output into Preview Image or Save Image.
  4. Queue. The request runs on Runware and comes back as a native IMAGE.

What's in the pack

  • One node per model, grouped Runware/<Modality>/<creator>. Widgets are the model's real parameters, with correct ranges, defaults, and dropdowns.
  • Native outputs: image, upscale, and background-removal return IMAGE; audio returns AUDIO; video returns VIDEO; 3D and other files save to your output folder and return a path; text returns a string.
  • Native inputs: reference and seed images are IMAGE, inpainting masks are MASK; audio, video, and document inputs take a URL, path, or UUID.
  • Builder nodes (Runware/Params) keep model nodes clean. Stackable features (LoRA, ControlNet, IP-Adapter, Embeddings, and more) each wire into a model's typed socket, and you chain the stackable ones to combine them.
  • Custom checkpoints: Runware/Custom models has a node per architecture (SDXL, SD 1.5, FLUX, Pony, and more) for community fine-tunes.
  • New models, day one: Runware (custom) takes any model AIR and taskType as JSON and returns raw JSON. Pair it with Runware Get to pull a field out of the response (e.g. 0.imageURL).
  • Run info on the title bar: each run shows its cost and, when a content check ran, the NSFW result (e.g. $0.00078 · NSFW: no).

The full guide covers every builder, custom models, parameter behavior, and troubleshooting.

Links