The Cloud Asset Drive for ComfyUI · Windows, macOS & Linux

Your ComfyUI runs, preserved.
Never lose a working setup again.

Lock the exact models, custom nodes, dependencies, and workflow behind a run that actually worked. Free up local SSD space without fear — bring missing models and red nodes back anytime, or revive your environment on a cloud GPU in minutes.

One active workflow slot included free for a full week. Safe local SSD cleanup. No card needed.

ComfyUI on Windows, macOS and Linux · Cloud restore verified on Linux + NVIDIA (RunPod, Vast.ai)
Check compatibility & limits →

THE RUN THAT WORKED YOUR NEST ALL FOUR BACK capture restore

The run it came from

ComfyUI
kohya_ss · early
LLaMA-Factory · early

The models inside

SDXL
FLUX.2 klein 4B
Wan 2.2

All logos, trademarks and registered trademarks are property of their respective owners.

Three steps

Capture it, move it, restore it.

01Capture

Shut the pod down without losing the setup.

The render finally works. One command records what made it work — the models, the custom nodes, the exact locked versions, the workflow — so the machine underneath becomes disposable again. If a package can't be installed on another machine, the pack says so and marks the nest.

02Move nothing running OBJECT STORAGE stored once · never sent twice

It waits in storage, not on a machine you rent.

Your first active workflow sits in your drive free for a full week — or permanently on a paid plan. Nothing is running, nothing bills by the hour, and a file you have already stored is never stored twice.

03Restore match

Back on the next machine — and checked.

One command on whatever you rent next. Every file is checked byte for byte on the way in and the dependencies are reinstalled at the versions you packed. Then it starts the app, runs the packed workflow once and tells you what passed, what didn't, and where the output is. Where it stands today: the files come back, checked.

One measured case · 9 October 2026

Saved on RunPod. Restored on a clean Vast.ai instance.

Packed onRunPodRTX A5000 · Canada · 7 Oct 2026
Restored onVast.aiRTX 3090 · Ukraine · 9 Oct 2026
What movedAnima StandardA community ComfyUI workflow · 5 custom-node packs · 3 model files · 164 Python packages · 5.67 GB nest

Restore, step by step · seconds

27checking this machine
76downloading
52putting files in place
293installing dependencies
25starting up
48test render
8 min 56 sdoor to door · restore command, start to last line
6 filescompared afterwards · byte for byte identical
1 pasteone command copied from the console · nothing typed after
renest restore · S5 test render Read-only
✓ Re-ran the packed recipe and it produced 1 image in 48s

One measured run, by us. Renting and booting the machine are not in the time.

See the full record →

What travels with your workflow

Open the nest: what comes back, and what stays behind.

Restoring for yourself brings back every stored file, each one checked.

Comes back 35 items · checked
Models, app, node packs, workflow, lock, escape hatch.
Mind on a hand-off 1 model file · 603.0 MB
Left out of a hand-off unless confirmed as passable.
Never stored
output/ temp/ input/ user/
IP-Adapter + CLIPSeg showcasenest · packed 11 Oct 2026 · 36 items · 7.96 GB
Models7.80 GBCustom nodes39.8 MBApp, workflow, lock, escape hatch113.8 MB
Name Type Size
v1-5-pruned-emaonly.safetensors
Checkpoint
4.27 GB
t2iadapter_depth_sd15v2.bin
ControlNet
309.5 MB
CLIP-ViT-H-14-laion2B-s32B-b79K.safetensors
CLIP model
2.53 GB
ip-adapter-plus_sd15.safetensors
Ipadapter
98.2 MB
model.safetensorsleft out of a hand-off
Checkpoint
603.0 MB

No licence from the author, so it does not travel in a hand-off: the recipient's restore names it and its source, fetches it from there and checks it against the nest.

ComfyUI_IPAdapter_plus
Custom node
281.0 KB
ComfyUI_essentials
Custom node
63.7 KB
comfyui_controlnet_aux
Custom node
39.4 MB
ComfyUI
App source
11.2 MB
config.json
Model setting
4.7 KB
merges.txt
Other file
524.6 KB
preprocessor_config.json
Model setting
380 B
special_tokens_map.json
Other file
472 B
tokenizer_config.json
Model setting
974 B
vocab.json
Other file
1.1 MB
refs/main
Other file
40 B
added_tokens.json
Other file
0 B
audio_tokenizer_config.json
Other file
0 B
chat_template.jinja
Other file
0 B
chat_template.json
Other file
0 B
processor_config.json
Other file
0 B
tokenizer.json
Other file
0 B
put_blueprints_here
Other file
0 B
merges.txt
Other file
524.6 KB
special_tokens_map.json
Other file
472 B
vocab.json
Other file
1.1 MB
legacy_directory_do_not_use.txt
Other file
0 B
put_font_files_here.txt
Other file
0 B
put_luts_files_here.txt
Other file
0 B
depth_anything_v2_vits.pth.lock
Other file
0 B
depth_anything_v2_vits.pth
Other file
99.2 MB
apache-2.0.txt
Licence text
10.3 KB
creativeml-openrail-m.txt
Licence text
14.4 KB
restore.sh
Other file
144.8 KB
af28f9b8b68c.json
Other file
4.3 KB
requirements.lock
Lockfile
4.6 KB
An example nest36 items · 7.96 GB

Two ways to start

Try a run we packed. Then pack your own.

01

Try a run we packed

Browse the starter nests

Claim a starter nest — a hand-off from Renest — and restore it on a machine you rent.

Image made with the FLUX.2 Klein 4B environmentFLUX.2 Klein 4B
Image made with the Z-Image-Turbo environmentZ-Image-Turbo
Frame from the Wan 2.2 TI2V 5B environmentWan 2.2 TI2V 5B
4 min 24 srestore start to image · RunPod RTX A5000 · 12.46 GB · one measured run
02

Then pack your own

Open the quick start

Get the run working first — then save it in one click from the ComfyUI sidebar panel (Windows, macOS, Linux) or with the CLI. Your first active workflow sits in your drive free for a full week.

Shell
$ uv tool install --upgrade renest$ renest pack --dir ./run --workflow workflow-api.json --out ./nests --dest hosted

Where the promise stops

Promised, tested, and not offered.

Promised
Your files come back, checkedByte for byte; a mismatch names the file.
Your dependencies come backNodes at the commit that ran, packages as packed.
Tested every restore
— not promised
The app startsStarted on every restore; you're told if it answered.
Output comes outThe packed workflow runs once; you see what passed.
Not offered
The same image on any hardwareOther hardware can give a different image. See the comparison
Getting an unfamiliar setup runningRenest keeps a run that already worked.
Public linksA hand-off is always private.

Source-available tool · Open format

Walk out with everything — one command, any time.

The format spec and the escape hatch are Apache-2.0; the renest tool's source is public, but the tool is not open-source software. Nothing about a nest depends on us still being here — your work is never hostage to us.

  • ✓Needs nothing of ours — one shell script and the commands your machine already has.
  • ✓Asks no server — it runs on its own, even if we are gone.
restore.sh the escape hatch

Every nest ships one. It brings back the files and the locked dependencies with zero Renest code.

bash restore.sh

Point it at your nest and a target folder. It leans on curl · jq · sha256sum · tar · uv — and nothing else. escape hatch

Hand-off

Pass a working run to someone. Nothing gets re-uploaded.

A hand-off is a private link: they claim the nest into their own drive and restore it on their own machine. A file whose licence doesn't let you pass it on does not travel: their restore names it and where it comes from, fetches it from its original source and checks it against the nest.

YOUR DRIVE THEIR DRIVE one nest stays yours hand-off link their copy no re-upload already stored — nothing of yours is uploaded again

Private link, on your terms

A hand-off is an unguessable private link, single-use by default. You choose how many people can claim it — up to a limit set by your plan — and how long it lives. It stops the moment the claims run out, the clock runs out, or you revoke it.

Never in a public index

Not just "no gallery": the preview page is never indexed by search engines. It shows a summary — name, size, how many files, dates — never the file names, never an image. Nothing for a stranger to stumble into — or a crawler to find.

Your original stays yours

Claiming copies the nest into their drive without consuming yours.

Pole to pole · ~71,000 km a year

Restored as far as the Arctic tern flies.

The Arctic tern makes the longest migration on Earth — pole to pole and back, every year. A nest travels the same way: packed once where the run lives, kept in storage while nothing runs, and rebuilt file for file on the next pod you rent.

One global network

Restore where the GPUs are.

A nest comes down from your drive onto whichever machine you rent — RunPod, vast.ai or your own — every file checked byte for byte.

North America · South America · Europe · South Asia · Southeast Asia · East Asia · Oceania

FAQ

Questions, answered plainly.

Why does my ComfyUI setup disappear when a RunPod pod stops, and how do I keep it?

A pod's container disk is wiped when it stops, and the custom nodes, their Python packages and downloaded models go with it. A network volume keeps the files inside one region of one provider; a nest keeps the files and a reinstallable record of the environment, and comes back on any machine you rent, every file checked.

Keep ComfyUI when a pod stops

I already use a network volume. Why would I need anything else?

If you always come back to the same region and the same kind of machine, you may not. A volume keeps files, not a working environment: the Python packages and system libraries a custom node needs are built for one machine and often break on the next one. A nest reinstalls them for the machine you landed on.

Network volume, new pod, still broken

Is this different from a ComfyUI-Manager snapshot or a Docker image?

Yes — each keeps a different layer. A Manager snapshot records which node repositories and commits you had and clones them again; it doesn't carry models. A Docker image freezes the system and Python layers, and gets huge with models baked in. A nest keeps the files themselves, the exact package versions to reinstall, and the system libraries the run loaded.

Volume, image, snapshot or nest

Will the same seed give the same image after I move to another GPU?

Not necessarily, and we don't promise it. The same GPU model with the same CPU brand gives the same image; other hardware gives small, consistent differences. What we guarantee is the files.

Same GPU, two images

What do I need on the rented machine?

A Linux machine with an NVIDIA GPU, and the renest command-line tool. The quickest start is the official base image — one template link on RunPod or vast.ai. Any other machine works too: the restore checks it before anything downloads and refuses a pairing that cannot work.

Renest on RunPod

The restore finished. Now what?

It prints where the output is, the workflow file, and the command that starts the app: renest start --dir ./run --listen 0.0.0.0. Open that port in your provider's panel and use your ComfyUI in the browser as usual.

After the restore

Can I hand a working setup to a client or teammate?

Yes, with a hand-off: an unguessable private link you send yourself — nothing is ever listed publicly. They preview what's inside, claim it into their own drive and restore it on a machine of their own. Receiving works on the free plan; sending needs a paid plan.

Give a working setup to someone

What happens to my files if Renest disappears?

They stay yours. The format is open (Apache-2.0) and every nest carries a plain restore.sh that brings its files and dependencies back from storage with ordinary tools and no Renest code. You can also download any version, any time.

The escape hatch

Turn today's success into a place you can return to.

Save your first active workflow to your drive free for a full week — or test how fast an environment revives with a verified setup. Free, no card.