Best viewed on a desktop or Mac. This page carries wide code blocks, terminal output and a 276 model reference table. On a phone you can read it, but you cannot set it up.

Repo preview · ships to GitHub next

A node canvas for AI, woven by your coding agent.

Wire 276 image, video, 3D and audio models into visual pipelines. Bring your own keys, pay providers at cost. Then point Claude Code, Codex or Cursor at it and watch the canvas build itself.

Apache-2.0Next.js 14 TypeScriptReact Flow MCP built in0 telemetry

Connect your IDE See the app

a real pipeline, really running · localhost:3005

276

AI models behind one key

node types on the canvas

pre-wired pipeline templates

MCP tools for your agent

providers, bring your own keys

databases, accounts, telemetry

Product tour

Five screens that explain the whole thing.

Every shot below is the actual app running locally. No mockups, no concept art.

Blank canvas with the three-step How to use Loometo card and 22 template chips.

first launch · empty state

01 · First launch

You're never staring at an empty grid.

Open Loometo and it teaches itself: a three-step explainer, every template one click away, and a search over all 57 nodes.

Three steps, on screen. Set up keys, drag nodes, hit Run. The explainer card covers the entire learning curve.
22 templates as chips. UGC Clip, Launch Kit, 3D Product Spin, Headshot Studio… click one and a pre-wired pipeline lands on the canvas.
Node search + grouped sidebar. Text & Prompts, Image, Video, Audio, 3D, Utility: every node one keystroke away.

A Text to Image node with its model dropdown open, showing Flux 2 Pro with a PICK badge and premium price tier, plus Nano Banana Pro and Seedream options with cost tiers.

text → image node · model picker open

02 · Model intelligence

Every model, priced and ranked, inside the node.

No tab-switching to compare models. The picker is searchable, cost-labeled and curated, and the parameters below it are schema-driven, so you can never set a value the model rejects.

★ PICK badges. Curated recommendations per category, so newcomers land on the right model first try.
$ / $$ / $$$ cost tiers on every entry (cheap, medium, premium) before you spend a cent.
Provider pills + routing. OpenAI · Gemini · Loometo pills up top; the "VIA MUAPI" tag shows exactly which route your call takes.
Schema-driven params. Aspect ratio, resolution, negative prompt: the dropdowns only offer values this exact model accepts.

The API Keys settings panel listing providers with masked key fields, FROM .ENV chips, Get key links and a Save Keys button.

settings → api keys · values masked

03 · Key management

Paste, Save, generate. Keys never leave your machine.

Hit Setup Keys once, paste, press Save. Every provider works immediately, no config files, no restart. Each row explains what it powers and links straight to where you get the key.

One required key. The rest are optional upgrades, the panel marks exactly which is which.
FROM .ENV chips. Using a .env.local file instead? Those keys are detected, badge-marked, and always win over pasted ones.
Masked with reveal toggles. Every field ships hidden; per-row "Get key" links jump to the provider's key page.

A Remove Background node with its info tooltip open, explaining what it does, with Takes in: Image and Outputs: Cutout.

remove background node · info open

04 · Node anatomy

Every node explains itself in plain English.

Shown here: Remove Background. Tap the (i) on any of the 57 nodes and it tells you what it does, what it takes in, and what comes out, before you spend anything.

Plain-English info. "Subject stays, background disappears. Perfect for cutout images." No jargon, no doc-hunting.
Typed handles. TAKES IN: Image → OUTPUTS: Cutout. Wires only connect where the types match, so mis-wiring is impossible.
Tap to run, per node. Run one node, a branch, or the whole graph. Outputs flow forward automatically.

The Director Formula node with eleven fields: subject, action, shot type, lens, three lighting facets, textures, composition, style reference and emotional vision, outputting a Director Prompt.

director formula node · 11 fields → one prompt

05 · The director's cut

Prompt like a film director, not a prompt engineer.

The Director Formula node encodes real cinematography discipline: SHOT + LENS + LIGHT + TEXTURE + COMPOSITION + STYLE REF + VISION. Fill the fields like a creative brief; it assembles a director-grade prompt that wires into any generator.

Real shot vocabulary. ECU → EWS shot codes, lens bands from 14mm ultra-wide to telephoto, each option annotated with the feeling it produces.
Lighting in three facets. Source, direction, quality: the same split a DoP uses, as three dropdowns.
Presets. Save a look you like and reapply it. The PRESET button banks the whole 11-field configuration.
Part of a prompting suite. Prompt Enhancer, Prompt Combiner, Image Describer and Image → JSON sit beside it: a full prompt-intelligence toolkit, in nodes.

The skills

57 nodes, eight studios, one canvas.

Mix them freely. A board can feed a video node, which feeds an upscaler, which feeds the output.

Reference Boards

Lock a character, product or place once, reuse it across every shot for consistency no single prompt can give you.

Character BoardProduct BoardLocation BoardMascot BoardCreature BoardB-Roll BoardReference SheetStoryboard

Generate

The core generators: text or image in, image or video out, on whichever of the 276 models you pick.

Text → ImageText → VideoImage → ImageImage → VideoVideo → VideoSpeech → Video

Edit & Retouch

Surgical image ops without leaving the canvas: remove, replace, relight, upscale, swap outfits and faces.

InpaintOutpaintObject RemoveFace SwapInter ChangeRelightBG RemoveUpscale

Mask & Composite

Type what to mask and get an alpha. Grow mattes, merge channels, composite layers, pull video mattes.

Mask by TextMask ExtractorMatte Grow/ShrinkMerge AlphaCompositorVideo MatteExtract Frame

Audio & Lip Sync

Voiceover, music and talking avatars: generate speech, clone a voice, transcribe, lip-sync it to video.

TTSVoice ClonerAudio TranscriberLip Sync

3D

Any image becomes a mesh. Orbit it inside the node, then export GLB for the web or a print-ready STL.

Image → 3DOrbit in-nodeGLB exportSTL export

Prompt & LLM

Prompting help built in: enhance a rough idea, apply the director formula, describe an image, extract structured JSON.

Prompt EnhancerDirector FormulaPrompt CombinerRun Any LLMImage DescriberImage → JSON

Import, Output & Utility

Assets in, results out. Paste or upload media, pull a URL or a Figma frame, compare variants, batch with the iterator.

Image/Video/Audio UploadPaste from clipboardURL ImportFigma ImportPainterCompareIteratorOutput

The catalog

276 models across 10 categories. The whole market, one schema.

The catalog lives in public/muapi-schema.json. Every selector reads its allowed values from there. New model out today? Add one schema entry and it appears in the app. Cost tiers $ cheap · $$ medium · $$$ premium show on every model before you run it.

ROUTE ADirect provider

For providers you hold a key to: Google (Imagen, Veo, Gemini), OpenAI (GPT-Image, DALL·E), Meshy (3D), ElevenLabs (voice). Loometo calls them straight: raw cost, zero middleman markup.

ROUTE BMuAPI reseller

One MUAPI_API_KEY unlocks the other ~250 models (Kling, Sora, Runway, Wan, Flux, Tripo, Suno and more) without signing up to a dozen labs. lib/api/router.ts picks the route per model, automatically.

Image → Video62 models

kling-v2.5-turbo-proveo3.1runwayseedance-2.0-omni-referencewan2.6minimax-hailuo-02higgsfield-doppixverse-v5vidu-q1hunyuanltx+51 more

Image → Image56 models

nano-banana-editflux-kontext-maxhiggsfield-soulimage-faceswapobject-eraserproduct-shotskin-enhancerghibli-styleideogram-characterupscaler+46 more

Text → Image47 models

flux-2-proflux-kontext-pronano-banana-progoogle-imagen4-ultragpt-image-2midjourney-v7qwen-imagebytedance-seedream-v4.5hidream+38 more

Text → Video44 models

openai-sora-2-proveo3.1kling-v2.1-masterwan2.5minimax-hailuoseedancevidupixverse+36 more

Video → Video27 models

topaz-video-upscaleluma-modify-videorunway-alephvideo-faceswapvideo-translatewatermark-removeranime-restyle+20 more

Audio → Video · Lip Sync11 models

sync-lipsyncinfinitetalkkling-avatarcreatifyveed+6 more

Text → Audio8 models

suno-create-musicminimax-speech-hdspeech-turbommaudio-v2voice-clone+3 more

Image → 3D9 models

tripo3d-v2.5meshy (direct)meshy-6 (via MuAPI)→ GLB / STL export+5 more

Text → Text · Training12 models

gpt5-minigpt5-nanostoryboardimage-to-textflux-dev-lorasdxl-lora+6 more

Bring your own keys

Six providers. One required. Paste it in the app and you're done.

The whole setup, in plain words: get a key from muapi.ai, open Loometo, hit Setup Keys (top right), paste it, press Save. That single key unlocks ~80% of the app. No config files, no restart, nothing else.

EASY WAYPaste in the app, works for all six

Setup Keys → paste → Save. The key is stored in your browser and sent only to your own localhost server, which forwards it to the provider. Works instantly for every provider below.

CONFIG WAY.env.local file, optional

Prefer keys in a file? Copy .env.example to .env.local, add the key, restart the dev server. Supported for MuAPI, Google and ElevenLabs. A file key always overrides a pasted one.

Provider What it powers Get your key Paste in-app .env.local

MuAPIREQUIRED Video, 3D, upscale, image edit, face swap, lip sync: the ~250-model reseller catalog muapi.ai ✓ works ✓ MUAPI_API_KEY

Google AIOPTIONAL Imagen 4, Veo 3, Gemini, called direct, bypassing reseller markup aistudio.google.com/apikey ✓ works ✓ GOOGLE_API_KEY

ElevenLabsOPTIONAL Text-to-speech + voice cloning (free tier available) elevenlabs.io → API keys ✓ works ✓ ELEVENLABS_API_KEY

OpenAIOPTIONAL GPT-Image / DALL·E, plus the LLM and prompt nodes platform.openai.com/api-keys ✓ works —

MeshyOPTIONAL Direct image → 3D, the cheapest 3D path meshy.ai/api/keys ✓ works —

FigmaOPTIONAL Pull frames and designs straight onto the canvas figma.com → access tokens ✓ works —

Where do keys actually live? Pasted keys stay in your browser's storage and travel only to localhost, your own machine, never to any Loometo server (there isn't one). Keys in .env.local never reach the browser at all.

Optional extra: set Cloudflare R2_* keys in .env.local and every generated asset is auto-copied to your own bucket. Provider URLs expire in 24 to 72 hours. This makes them permanent.

The part no other node canvas has

Connect your coding agent. Watch the canvas build itself.

Loometo ships a zero-dependency MCP server (mcp/loometo-mcp.mjs, Node 18+) wrapping its HTTP canvas bridge. Your agent reads the graph, adds nodes, wires them and fires runs, live, in your browser tab.

claude code · loometo

you › build me a UGC pipeline for a street-food ad

→ snapshot() reads empty canvas

→ add_node(character-board) ✓ char-board-1

→ add_node(image-to-video, kling-v2.5) ✓ i2v-1

→ connect(char-board-1 → i2v-1) ✓

→ set_param(i2v-1, aspect 9:16) ✓

→ run_all() ✓ generating…

canvas updated live, 4 nodes wired, running.

The 11 tools your agent gets

snapshot	Read the canvas (nodes + edges), call first
add_node	Add a node by type → returns its id
connect	Wire source → target, by id or card name
set_param	Set prompt, model, aspect ratio via a patch
run_node / run_all	Fire one node or the whole graph
delete_node / clear	Remove a node / wipe the canvas
select / pulse	Highlight or flash nodes for the user
list_models	Valid model slugs + params per category

Start Loometo, leave the tab open

That browser tab is what updates live while the agent works. First time here? Set your keys before the agent runs anything: Setup Keys → paste your MuAPI key → Save (see the API keys section), otherwise every run the agent fires will fail with a missing-key error.

terminal

npm run dev   # serves on http://localhost:3005

Register the MCP server in your agent

Use the absolute path to mcp/loometo-mcp.mjs. Set LOOMETO_URL only if you changed the port. Pick your client:

CCClaude Code

# one-liner, or edit ~/.claude.json

claude mcp add loometo \
  -- node \
  /abs/path/mcp/loometo-mcp.mjs

CXCodex

~/.codex/config.toml

[mcp_servers.loometo]
command = "node"
args = ["/abs/path/mcp/
  loometo-mcp.mjs"]

CUCursor / Cline

MCP settings → add server

{ "command": "node",
  "args": ["/abs/path/
    mcp/loometo-mcp.mjs"] }

the full block · .mcp.json or ~/.claude.json

"mcpServers": {
  "loometo": {
    "command": "node",
    "args": ["/abs/path/to/loometo/mcp/loometo-mcp.mjs"],
    "env": { "LOOMETO_URL": "http://localhost:3005" }
  }
}

Just ask it to build

Reload your agent so it picks up the server, then talk in plain English. Nodes are addressable by the name on the card, you can literally say "run char-board-1".

you › Look at the canvas, then add a Character Board, a Location Board, three Image→Video nodes on kling-v2.5, and wire them into Outputs.

you › Load the "Full UGC Campaign" template, swap every video model to veo3.1, set them all to 9:16.

No server running? Every tool returns a friendly "start Loometo first" instead of an error dump.

Templates

22 production-tested pipelines, one click each.

Every template is a pre-wired canvas: models pre-picked, prompts pre-filled with director-formula briefs you edit in place.

UGC ClipPerson + place → video

Product ShotsOne photo → angle board + cutout

Link → CarouselPaste a link, post a carousel

Launch KitOne hero → every channel

3D Product SpinPhoto → spinnable 3D

Style RemixPhoto + idea → upscaled art

Face SwapTwo photos → swapped

Video RestyleClip + idea → new look

Talking AvatarPhoto + script → talking video

Full UGC CampaignCharacter + place + product → 3 clips

Carousel from IdeaTopic → carousel

Voiceover ReelScript + face → talking clip

Animated LogoLogo → motion sting

Sticker PackPhoto → 4 die-cut stickers

Background SwapCut out → drop in a scene

Relight ProductPhone photo → studio light

Virtual Try-OnPerson + outfit idea → look

Headshot StudioSelfie → pro headshots

Photo RestoreOld / blurry → sharp

Object RemoverDelete anything from a photo

Ad Variant FactoryOne hero → 6 ad angles

Blog HeroArticle link → hero image

Quick start

Running in four lines.

terminal

# clone, install, go
git clone https://github.com/…/loometo.git
cd loometo && npm install
npm run dev                 # → localhost:3005
# then in the app: Setup Keys → paste MuAPI key → Save

No database to provision, no Docker, no accounts, no config files. Node 18+ and one MuAPI key pasted in the app is the entire prerequisite list.

In the box

Repo layout.

dinimiciuil-labs/loometo

app/
  api/proxy/     # key-holding provider proxies
  api/canvas/    # the agent bridge (HTTP + SSE)
  demo/[slug]/   # public read-only showcase
components/
  FlowCanvas.tsx  # the React Flow canvas
  nodes/         # all 57 node types
  ui/ sidebar/   # pills, pickers, panels
lib/
  api/           # dispatch, router, capabilities
  flow/          # templates, propagate, GLB→STL
  server/        # localhost + CSRF guard
mcp/            # zero-dep MCP server
public/
  muapi-schema.json # the 276-model catalog
LICENSE  NOTICE  TRADEMARK.md

Security posture

Local-first, audited before release.

Loometo is built to run on your machine. The public repo shipped only after a security pass. These guards are already in the code.

Localhost-only API routes

Every proxy and canvas-bridge route rejects non-local requests with a 403. Nothing answers the outside world.

CSRF-hardened bridge

Mutating endpoints require application/json, forcing a failing preflight on any cross-origin attempt from a malicious page.

Zero telemetry

No analytics, no accounts, no database, no phone-home. Your prompts and assets stay on your machine (or your own R2 bucket).

Keys held server-side

.env.local keys are injected by server proxies and never reach the browser. In-app keys never leave your browser.

One honest caveat, stated in the README too: the dev server has no auth by design. Don't port-forward it to the public internet without adding your own auth layer.

Complete reference

Every node. Every model. Documented.

Pulled straight from the codebase: the same plain-English explanations behind each node's (i) button, and the full 276-entry model catalog with each model's extra parameters. Click a group to expand it.

All 57 nodes, in plain English

Text & Prompts8 nodes

Node	What it does, from the (i) button
Prompt	Type what you want to create here. Like telling the AI 'make me a sunset on a beach'. It's the starting point of everything.
Prompt Enhancer	Your prompt goes in, a better prompt comes out. The AI rewrites it with more detail so your image or video looks amazing.
Audio Transcriber	Listens to your uploaded audio and writes out the spoken script. Feed the transcript into a Prompt Combiner so the video model knows what is being said, not just lip shapes.
Prompt Combiner	Connect two prompts and this merges them into one. Useful when you have different ideas you want to blend together.
Director Formula	Fill the 7 fields like a creative director writing a brief: subject + action, shot type, lens, lighting (3 facets), composition, style ref, vision. The node assembles a director-grade prompt that wires into any image generator. Vocabulary taken verbatim from cinematography sources.
Run Any LLM	Talk to any AI brain (ChatGPT, Gemini, Claude) directly inside your workflow. Ask questions, get descriptions, or have it write prompts.
Image Describer	Give it a photo and it tells you exactly what's in it as a detailed prompt. Perfect for recreating a style.
Image → JSON	Vision LLM produces a strict JSON description of one or several images of the same subject. Wire optional Extra Instructions (positive direction: 'include hex codes', 'list every logo separately') and Negative (avoid: 'don't speculate about brands not visible'). Output JSON wires into any text input for downstream identity lock.

Audio & Voice5 nodes

Node	What it does, from the (i) button
Text → Speech	Turns a script into a spoken audio clip. Wire a Prompt with your script in, get audio out. For ElevenLabs free tier, wire a Voice Cloner in (library voices are paid-only). Pipe that into Lip Sync to make a video character say it.
Voice Cloner	Wire in a 30+ second audio sample (your voice, a podcast clip, anything clean). Give it a name. Click Clone. ElevenLabs returns a voice_id you can wire into Text → Speech for unlimited future TTS in that voice. Free tier compatible.
Lip Sync	Upload a video of a person and an audio clip. It animates their lips to match the audio.
Speech → Video	Take a portrait photo, add a voice recording, and get a video of that person speaking. Perfect for AI avatars.
Upload Audio	Pick an audio file (mp3, wav). Feed it into Lip Sync or Speech → Video.

Video7 nodes

Node	What it does, from the (i) button
Video Describer	Same as Image Describer but for videos. Watches your video and writes out what's happening so you can recreate or remix it.
Text → Video	Type what you want to see and it creates a video. The dropdown shows only what each model truly supports.
Image → Video	Animate a single photo (First Frame), wire First + Last for transition (Veo 3.1), or wire multiple references for reference-conditioned models. Ref handle count is set by the chosen model: Sora 2 / Veo 3 base = 1, Pixverse = 2, Veo 3.1 Reference = 3, Wan 2.1 Reference = 5, Vidu Q1/Q2 + Kling O1 Reference = 7, Seedance Omni = 9. Address each in the prompt as @image1..@imageN.
Video → Video	Transform a video's look completely. 'Make this look like anime' or 'make it look like 1970s film.' Same motion, new style.
Video Upscale	Makes blurry or low-quality videos look sharp and crisp.
Video Matte	Removes the background from every frame of a video automatically. Tracks your subject, like a green screen without the green screen.
Upload Video	Pick a video file (mp4, mov, webm). Feed it into Video → Video, Lip Sync, or other video tools.

3D1 nodes

Node	What it does, from the (i) button
Image → 3D	Generate a 3D mesh from one image, multiple angle photos, or a text prompt. Tripo3D H31 ($0.20-$0.30) is the cheapest; Meshy 6 ($0.50) is most detailed. Drag the mesh to rotate. After the GLB loads, a 2D snapshot is auto-captured so downstream image nodes have a usable image to consume; download serves the .glb.

Import, Output & Utility8 nodes

Node	What it does, from the (i) button
Extract Frame	Grab any single frame from a video and save it as a still image. Pick the exact moment you want, frame by frame.
Painter	Brush directly on the canvas. Paint MASKS for Inpaint / Object Remove, or sketch overlays that flow downstream as images.
Import from Figma	Paste a Figma frame URL. Right-click any frame in Figma → Copy link to selection. Hit Run to pull it in as an image.
Upload Image	Pick an image from your computer. It becomes the starting point for any image workflow.
Import from URL	Paste any website link. Free mode extracts the readable text directly ($0). Google mode uses Gemini to clean the text AND recover the page's images and logo. Tap the recovered assets to pick which ones flow into your carousel as references. A tick marks the chosen ones.
Compare	Put two images side by side with a draggable slider. Perfect for before/after.
Iterator	Run the same process on multiple images at once. Instead of generating 10 images one by one, the Iterator does them in a batch.
Output	The finish line. Connect image, video, and/or audio here to preview and download each. Pair Veo video + TTS audio here, then combine in your editor (CapCut / DaVinci).

Image & Boards28 nodes

Node	What it does, from the (i) button
Text → Image	Type a description and it draws a picture. Models that accept a STYLE REFERENCE image (Nano Banana, Flux Kontext, GPT-Image) expose a second Style Ref handle. Pure T2I models hide it.
Image → Image	Give it a photo + a description and it transforms the photo. Like 'make this look like an oil painting.'
Upscale	Makes small or blurry images bigger and sharper. It actually invents new detail. Turn a tiny image into a crisp 4K masterpiece.
Remove Background	Cuts out the background of any photo automatically. Subject stays, background disappears. Perfect for cutout images.
Relight	Changes where the light is coming from in a photo. Make daytime look like sunset, or add dramatic studio lighting, without a studio.
Inpaint	Paint over something you don't want and the AI fills it in naturally. Remove people, erase logos. It blends in seamlessly.
Outpaint	Makes your image wider or taller by generating what would be outside the frame. Like zooming out. The AI imagines what's beyond the edges.
Face Swap	Put your face (or anyone's face) onto a different body or scene. Handles skin tone, lighting and angle automatically.
Object Remove	Point at anything and poof, it's gone. The AI fills in what the background looks like. Remove tourists, wires, anything.
Compositor	Visual layer editor. Wire a SCENE + up to 4 SUBJECTS, then drag / resize / reorder them in the polaroid. Two output modes: FLAT (exact pixel composite, no AI) or AI BLEND (sends the arrangement to Nano Banana Pro Edit which re-renders with matched lighting and perspective).
Reference Sheet	Feed up to 6 photos of the SAME thing (a person, a place, a product) and one prompt. The model uses them all as identity / location anchors. Better than asking one model to 'imagine four angles' from a single photo.
Character Board	Wire 1-5 photos of a person AND optionally a JSON spec from Image→JSON. Generates a studio model board (8 head expressions, 4 full-body angles, hand/eye/shoulder close-ups, fabric and skin-tone swatches) with anti-fake realism + director-formula lighting baked in. Feed the output into Image→JSON downstream to lock identity for compositors.
Location Board	Wire 1-5 photos of a place AND optionally a JSON spec. Generates a location reference board: exterior angles, time-of-day variants (golden/midday/blue/night), signage close-ups, brand swatches. Director-formula lens + 4-facet lighting baked per panel. Feed the output into Image→JSON downstream.
Product Board	Wire 1-5 photos of a product AND optionally a JSON spec. Generates a spec-board: hero angles, in-hand shots, brand-mark and material close-ups, colourway swatches. Director-formula lens + 4-facet softbox lighting baked per panel. Feed the output into Image→JSON downstream.
B-roll Board	Wire 1-5 photos of your scene / subject / product. Generates 6-9 cutaway-style frames sharing one colour grade and lighting era: hands, atmosphere details, environment beats, lifestyle moments. Use for video edit asset library or campaign supporting shots.
Storyboard	Wire 1-5 photos of your character / scene / setting + an extra-direction line describing the story beat-by-beat. Generates 6-9 numbered panels, each labelled with shot type (ECU / CU / MS / FS / LS) and a one-line action. Photoreal hero frames at pre-vis quality.
Mascot Board	Wire 1-5 photos / concept art of a brand mascot. Generates a mascot bible: 4 orthographic views (front / 3-4 / side / back), 6 expressions, 4-5 poses, brand palette swatches, face + hand detail callouts, scale reference next to a human. Locks the mascot for consistent reuse across campaigns.
Creature Board	Wire 1-5 photos / concept art of a creature, monster, or designed animal. Generates a creature concept-art bible: 5 orthographic views (front / 3-4 / side / back / top-down), face / limb / signature-feature / texture / joint detail callouts, material swatches (skin / fur / scales / feathers), threat-display + resting + hunting pose silhouettes, scale reference next to a human.
Inter Change	Virtually try on different outfits. Describe the clothing and the AI puts it on the person. Great for fashion and e-commerce.
Crop	Cut your image or video to a specific size. Choose 16:9 for YouTube, 9:16 for TikTok, or 1:1 for Instagram. Instant, no AI needed.
Blur	Makes things fuzzy. Blur backgrounds, censor things, or create a dreamy soft-focus effect. Slide to control how blurry.
Levels	Make your image brighter, darker, or more contrasty. Like the basic sliders in your phone's photo editor but in your workflow.
Invert	Flips all colors to their opposite, like a photo negative. Black becomes white. Very useful for flipping masks.
Mask Extractor	Click on objects in your image and it automatically traces around them. Magic scissors that cut out exactly what you want.
Mask by Text	Describe what to select, 'the person's hair' or 'the car', and it draws the selection automatically. No clicking needed.
Matte Grow/Shrink	Makes your selection slightly bigger or smaller. Use this to clean up rough edges.
Merge Alpha	Takes your image and your mask and combines them. The masked area becomes transparent. How you get cutout images with no background.
Carousel Board	Turns any text (wire in an Import from URL, or paste directly) into a ready-to-post social carousel: hook slide, point slides, CTA slide. Wire a Logo in to brand every slide and a Reference Image for the hook background. Finish 'Flat' renders locally for free; 'AI Enhance' passes each slide through Nano Banana Pro Edit (~$0.12/slide). One output handle per slide.

The full model index, all 276

Image to Video62 models

Model slug	Variant	Extra parameters
ai-video-effects	AI Video Effects effects	name, aspect_ratio, resolution, quality
motion-controls	Motion Controls effects	name, aspect_ratio, resolution, quality
vfx	VFX effects	name, aspect_ratio, resolution, quality
veo3-image-to-video	Image to Video veo	images_list, aspect_ratio
veo3-fast-image-to-video	Image to Video [Fast] veo	images_list, aspect_ratio
runway-image-to-video	Image to Video runway	aspect_ratio, resolution, duration
wan2.1-image-to-video	Image to Video wan2.1	aspect_ratio, resolution, quality, duration
midjourney-v7-image-to-video	Image to Video midjourney	aspect_ratio, resolution, num_videos, variety
hunyuan-image-to-video	Image to Video hunyuan	aspect_ratio
seedance-2.0-omni-reference	Seedance 2.0 Omni Reference bytedance	images_list, aspect_ratio, quality, duration
seedance-2-vip-omni-reference-fast	Seedance 2 VIP Omni Reference Fast bytedance	images_list, aspect_ratio, duration
seedance-lite-i2v	Lite Image to Video bytedance	last_image, resolution, duration, camera_fixed
seedance-pro-i2v	Pro Image to Video bytedance	resolution, duration, camera_fixed
kling-v2.1-master-i2v	Master Image to Video kling-v2.1	aspect_ratio, duration
kling-v2.1-standard-i2v	Standard Image to Video kling-v2.1	aspect_ratio, duration
kling-v2.1-pro-i2v	Pro Image to Video kling-v2.1	last_image, aspect_ratio, duration
wan2.2-image-to-video	Image to Video wan2.2	last_image, aspect_ratio, resolution, quality
runway-act-two-i2v	Act 2 Image to Video runway	reference_video_url, aspect_ratio
pixverse-v4.5-i2v	Image to Video pixverse-v4.5	images_list, aspect_ratio, resolution, duration
vidu-v2.0-i2v	Image to Video vidu-v2	images_list, aspect_ratio, resolution, duration
vidu-q1-reference	Reference I2V vidu-q1	images_list, aspect_ratio
minimax-hailuo-02-standard-i2v	Standard I2V minimax-2	end_image_url, duration, resolution
minimax-hailuo-02-pro-i2v	Pro I2V minimax-2	end_image_url, duration, resolution
video-effects	Video Effects effects	name
pixverse-v5-i2v	Image to Video pixverse-v5	images_list, aspect_ratio, resolution, duration
seedance-lite-reference-video	Lite Reference to Video bytedance	images_list, resolution, duration
wan2.1-reference-video	Reference to Video wan2.1	images_list, resolution, aspect_ratio, duration
kling-v2.5-turbo-pro-i2v	Pro Image to Video kling-v2.5	duration
wan2.5-image-to-video	Image to Video wan2.5	resolution, duration
wan2.5-image-to-video-fast	Image to Video (Fast) wan2.5	resolution, duration
openai-sora-2-image-to-video	Sora 2 Image to Video sora	images_list, aspect_ratio, duration, remove_watermark
ovi-image-to-video	Image to Video ovi	prompt only
openai-sora-2-pro-image-to-video	Sora 2 Pro Image to Video sora	images_list, aspect_ratio, duration, resolution
leonardoai-motion-2.0	Motion 2.0 I2V leonardoai	aspect_ratio
higgsfield-dop-image-to-video	Image to Video higgsfield	last_image, motion, strength, options
veo3.1-image-to-video	Image to Video veo3.1	last_image, aspect_ratio, duration, resolution
veo3.1-fast-image-to-video	Image to Video [Fast] veo3.1	last_image, aspect_ratio, duration, resolution
veo3.1-reference-to-video	Reference to Video veo3.1	images_list, resolution, duration, generate_audio
seedance-pro-i2v-fast	Pro Image to Video Fast bytedance	resolution, duration, camera_fixed
ltx-2-pro-image-to-video	Pro Image to Video ltx	duration, generate_audio
ltx-2-fast-image-to-video	Fast Image to Video ltx	duration, generate_audio
vidu-q2-reference	Reference I2V vidu-q2	images_list, resolution, aspect_ratio, duration
vidu-q2-turbo-start-end-video	Turbo I2V vidu-q2	last_image, resolution, duration, bgm
vidu-q2-pro-start-end-video	Pro I2V vidu-q2	last_image, resolution, duration, bgm
minimax-hailuo-2.3-pro-i2v	Pro I2V minimax-2.3	resolution
minimax-hailuo-2.3-standard-i2v	Standard I2V minimax-2.3	duration
minimax-hailuo-2.3-fast	Fast I2V minimax-2.3	duration, go_fast
kling-v2.5-turbo-std-i2v	Standard Image to Video kling-v2.5	duration
grok-imagine-image-to-video	Image to Video grok	images_list, mode, duration
kling-o1-image-to-video	Image to Video [Pro] kling-o1	last_image, aspect_ratio, duration
kling-o1-reference-to-video	Reference to Video [Pro] kling-o1	images_list, aspect_ratio, duration, keep_original_sound
kling-v2.6-pro-i2v	Image to Video kling-v2.6	duration, sound
pixverse-v5.5-i2v	Image to Video pixverse-v5.5	images_list, style, thinking, aspect_ratio
wan2.2-spicy-image-to-video	Spicy Image to Video wan2.2	resolution, duration
wan2.6-image-to-video	Image to Video wan2.6	resolution, duration, shot_type
kling-o1-standard-image-to-video	Image to Video [Standard] kling-o1	last_image, duration
kling-o1-standard-reference-to-video	Reference to Video [Standard] kling-o1	images_list, aspect_ratio, duration
seedance-v1.5-pro-i2v	Image to Video seedance-v1.5-pro	last_image, aspect_ratio, resolution, duration
seedance-v1.5-pro-i2v-fast	Image to Video [Fast] seedance-v1.5-pro	last_image, aspect_ratio, resolution, duration
ltx-2-19b-image-to-video	Std Image to Video ltx	resolution, duration
kling-v3.0-pro-image-to-video	Image to Video [Pro] kling-v3.0	last_image, duration, generate_audio
kling-v3.0-standard-image-to-video	Image to Video [standard] kling-v3.0	last_image, duration, generate_audio

Text to Video44 models

Model slug	Variant	Extra parameters
veo3-text-to-video	Text to Video veo	aspect_ratio
veo3-fast-text-to-video	Text to Video [Fast] veo	aspect_ratio
runway-text-to-video	Text to Video runway	aspect_ratio, resolution, duration
wan2.1-text-to-video	Text to Video wan2.1	aspect_ratio, resolution, quality, duration
hunyuan-text-to-video	Text to Video hunyuan	aspect_ratio
hunyuan-fast-text-to-video	Fast Text to Video hunyuan	aspect_ratio
seedance-lite-t2v	Lite Text to Video bytedance	aspect_ratio, resolution, duration, camera_fixed
seedance-pro-t2v	Pro Text to Video bytedance	aspect_ratio, resolution, duration, camera_fixed
kling-v2.1-master-t2v	Master Text to Video kling-v2.1	aspect_ratio, duration
wan2.2-text-to-video	Text to Video wan2.2	aspect_ratio, resolution, quality, duration
pixverse-v4.5-t2v	Text to Video pixverse-v4.5	aspect_ratio, resolution, duration
vidu-v2.0-t2v	Text to Video vidu-v2	aspect_ratio, resolution, duration
wan2.2-5b-fast-t2v	Fast Text to Video wan2.2	aspect_ratio, resolution
minimax-hailuo-02-standard-t2v	Standard T2V minimax-2	duration, resolution
minimax-hailuo-02-pro-t2v	Pro T2V minimax-2	duration, resolution
pixverse-v5-t2v	Text to Video pixverse-v5	aspect_ratio, resolution, duration
kling-v2.5-turbo-pro-t2v	Pro Text to Video kling-v2.5	aspect_ratio, duration
wan2.5-text-to-video	Text to Video wan2.5	aspect_ratio, resolution, duration
wan2.5-text-to-video-fast	Text to Video (Fast) wan2.5	aspect_ratio, resolution, duration
openai-sora	Sora Text to Video sora	aspect_ratio, resolution
openai-sora-2-text-to-video	Sora 2 Text to Video sora	aspect_ratio, duration, remove_watermark
ovi-text-to-video	Text to Video ovi	aspect_ratio
openai-sora-2-pro-text-to-video	Sora 2 Pro Text to Video sora	aspect_ratio, duration, resolution, remove_watermark
veo3.1-text-to-video	Text to Video veo3.1	aspect_ratio, duration, resolution
veo3.1-fast-text-to-video	Text to Video [Fast] veo3.1	aspect_ratio, duration, resolution
openai-sora-2-pro-storyboard	Sora 2 Pro Storyboard sora	shots, duration, images_list, aspect_ratio
veo3.1-extend-video	Extend Video veo3.1	request_id
seedance-pro-t2v-fast	Pro Text to Video Fast bytedance	resolution, duration, aspect_ratio, camera_fixed
ltx-2-pro-text-to-video	Pro Text to Video ltx	duration, generate_audio
ltx-2-fast-text-to-video	Fast Text to Video ltx	duration, generate_audio
minimax-hailuo-2.3-pro-t2v	Pro T2V minimax-2.3	resolution
minimax-hailuo-2.3-standard-t2v	Standard T2V minimax-2.3	duration
grok-imagine-text-to-video	Text to Video grok	aspect_ratio, mode, duration
kling-o1-text-to-video	Text to Video [Pro] kling-o1	aspect_ratio, duration
kling-v2.6-pro-t2v	Text to Video kling-v2.6	aspect_ratio, duration, sound
pixverse-v5.5-t2v	Text to Video pixverse-v5.5	style, thinking, aspect_ratio, resolution
wan2.6-text-to-video	Text to Video wan2.6	aspect_ratio, resolution, duration, shot_type
seedance-v1.5-pro-t2v	Text to Video seedance-v1.5-pro	aspect_ratio, resolution, duration, generate_audio
seedance-v1.5-pro-t2v-fast	Text to Video [Fast] seedance-v1.5-pro	aspect_ratio, resolution, duration, generate_audio
ltx-2-19b-text-to-video	Std Text to Video ltx	aspect_ratio, resolution, duration
veo3.1-4k-video	4k Video veo3.1	request_id
kling-v3.0-pro-text-to-video	Text to Video [Pro] kling-v3.0	aspect_ratio, duration, generate_audio
kling-v3.0-standard-text-to-video	Text to Video [standard] kling-v3.0	aspect_ratio, duration, generate_audio
seedance-v2.0-t2v	Seedance 2.0 bytedance	aspect_ratio, duration, quality

Image to Image56 models

Model slug	Variant	Extra parameters
ai-image-upscaler	Image Upscaler tools	prompt only
ai-image-face-swap	Image Faceswap tools	swap_url, target_index
ai-dress-change	Dress Change tools	model_image_url, garment_image_url
ai-background-remover	Background Remover tools	prompt only
ai-product-shot	Product Shot tools	scene_description
ai-skin-enhancer	Skin Enhancer tools	prompt only
ai-color-photo	Color Photo tools	prompt only
flux-kontext-dev-i2i	Kontext Dev I2I kontext	images_list, aspect_ratio, num_images
ai-product-photography	Product Photography tools	person_image_url, product_image_url
ai-ghibli-style	Ghibli Style tools	prompt only
ai-image-extension	Image Extension tools	prompt only
ai-object-eraser	Object Eraser tools	mask_image_url
flux-kontext-pro-i2i	Kontext Pro I2I kontext	images_list, aspect_ratio
flux-kontext-max-i2i	Kontext Max I2I kontext	images_list, aspect_ratio
gpt4o-image-to-image	Image to Image gpt	images_list, aspect_ratio, num_images
gpt4o-edit	Edit Image gpt	mask_image_url, aspect_ratio, num_images
midjourney-v7-image-to-image	Image to Image midjourney	speed, aspect_ratio, variety, stylization
gpt-image-2-image-to-image	GPT Image 2 (Image to Image) openai	images_list, aspect_ratio, resolution, quality
bytedance-seededit-v3	Edit Image v3 seedream	prompt only
midjourney-v7-style-reference	Style Reference midjourney	speed, aspect_ratio, variety, stylization
midjourney-v7-omni-reference	Omni Reference midjourney	speed, aspect_ratio, weight, variety
minimax-image-01-subject-reference	Subject Reference minimax	aspect_ratio, num_images
ideogram-character	Character ideogram	render_speed, style, aspect_ratio, num_images
flux-pulid	Pulid Image to Image flux	aspect_ratio
qwen-image-edit	Edit Image qwen	aspect_ratio
image-effects	Image Effects effects	name
nano-banana-edit	Edit Image nano	images_list, aspect_ratio
ideogram-v3-reframe	v3 Reframe ideogram	aspect_ratio, render_speed, style, num_images
bytedance-seedream-edit-v4	Edit Image v4 seedream	images_list, aspect_ratio, resolution, num_images
nano-banana-effects	Image Effects nano	name, aspect_ratio
flux-kontext-effects	Image Effects kontext	name
flux-redux	Redux Image to Image flux	aspect_ratio, num_images
qwen-image-edit-plus	Edit Image Plus qwen	images_list, width, height
wan2.5-image-edit	Edit Image wan2.5	images_list, width, height
higgsfield-soul-image-to-image	Image to Image higgsfield	style, aspect_ratio, strength, quality
reve-image-edit	Edit Image reve	prompt only
topaz-image-upscale	Image Upscale topaz	upscale_factor
seedvr2-image-upscale	Image Upscale seedvr2	resolution
qwen-image-edit-plus-lora	Edit Image Plus Lora qwen	images_list, rotate_right_left, move_forward, vertical_angle
nano-banana-pro-edit	Pro Edit Image nano	images_list, aspect_ratio, resolution
image-passthrough	Image to Image image	make_input
kling-o1-edit-image	Edit Image [Pro] kling-o1	images_list, aspect_ratio, resolution
flux-2-dev-edit	Edit Image [Dev] flux-2	images_list, width, height
flux-2-flex-edit	Edit Image [Flex] flux-2	images_list, aspect_ratio, resolution
flux-2-pro-edit	Edit Image [Pro] flux-2	images_list, aspect_ratio, resolution
vidu-q2-reference-to-image	Reference to Image vidu-q2	images_list, aspect_ratio, resolution
bytedance-seedream-v4.5-edit	Edit Image seedream-v45	images_list, aspect_ratio, quality
qwen-image-edit-2511	Edit Image 2511 qwen	images_list, width, height
wan2.6-image-edit	Edit Image wan2.6	images_list
qwen-text-to-image-2512	Text to Image 2512 qwen	width, height
gpt-image-1.5-edit	Image to Image gpt-1.5	images_list, aspect_ratio, quality
grok-imagine-image-to-image	Image to Image grok	prompt only
Api Node	Image to Image wavespeed	model_url, api_key, params
flux-2-klein-4b-edit	Edit Image [Klein 4B] flux-2	images_list, aspect_ratio
flux-2-klein-9b-edit	Edit Image [Klein 9B] flux-2	images_list, aspect_ratio
add-image-watermark	Add Image Watermark watermark	watermark_image_url, position, opacity, scale

Video to Video27 models

Model slug	Variant	Extra parameters
ai-video-face-swap	Video Faceswap tools	target_gender, target_index
mmaudio-v2-video-to-video	v2 Video to Video mmaudio	duration
runway-act-two-v2v	Act 2 Video to Video runway	reference_video_url, aspect_ratio
runway-aleph-v2v	Aleph Video to Video runway	aspect_ratio
luma-modify-video	Modify V2V luma	prompt only
luma-flash-reframe	Flash Reframe V2V luma	aspect_ratio, duration
ai-dance-effects	AI Dance Effects effects	resolution
infinitetalk-video-to-video	Audio to Video infinite-talk	resolution
ai-video-upscaler	Video Upscaler tools	resolution, copy_audio
wan2.2-edit-video	Edit Video wan2.2	resolution
heygen-video-translate	Video Translate tools	language
wan2.2-animate	Anime Video wan2.2	mode, resolution
topaz-video-upscale	Video Upscale topaz	upscale_factor
ai-video-upscaler-pro	Video Upscaler Pro tools	resolution
video-watermark-remover	Watermark Remover tools	prompt only
remix-video	Remix Video tools	aspect_ratio
video-passthrough	Video to Video video	make_input
kling-o1-video-edit	Edit Video [Pro] kling-o1	images_list, aspect_ratio, keep_original_sound
kling-o1-video-edit-fast	Edit Video Fast [Pro] kling-o1	images_list, aspect_ratio, keep_original_sound
wan2.2-spicy-video-extend	Spicy Video Extend wan2.2	resolution, duration
kling-o1-standard-video-edit	Edit Video [Standard] kling-o1	images_list, keep_original_sound
kling-v2.6-pro-motion-control	Pro Motion Control kling-v2.6	prompt only
seedance-v1.5-pro-video-extend	Video Extend seedance-v1.5-pro	resolution, duration, generate_audio, camera_fixed
seedance-v1.5-pro-video-extend-fast	Video Extend [Fast] seedance-v1.5-pro	resolution, duration, generate_audio, camera_fixed
kling-v2.6-std-motion-control	Std Motion Control kling-v2.6	prompt only
add-video-watermark	Add Video Watermark watermark	watermark_image_url, position, opacity, scale
ai-clipping	AI Clipping video	num_highlights, aspect_ratio, return_coordinates_only

Text to Audio8 models

Model slug	Variant	Extra parameters
mmaudio-v2-text-to-audio	v2 Text to Audio mmaudio	duration
suno-create-music	Create Music suno	style, model, instrumental, negative_tags
suno-remix-music	Remix Music suno	style, model, instrumental, negative_tags
suno-extend-music	Extend Music suno	style, model, continue_at, instrumental
minimax-voice-clone	Voice Clone minimax-2.3	custom_voice_id, model, need_noise_reduction, need_volume_normalization
minimax-speech-2.6-hd	Speech HD minimax-2.6	voice_id, speed, volume, pitch
minimax-speech-2.6-turbo	Speech Turbo minimax-2.6	voice_id, speed, volume, pitch
audio-passthrough	Text to Audio audio	make_input

Text to Image47 models

Model slug	Variant	Extra parameters
flux-dev	Dev flux	width, height, num_images
flux-kontext-dev-t2i	Kontext Dev T2I kontext	aspect_ratio, num_images
hidream-i1-fast	Fast hidream	width, height, num_images
hidream-i1-dev	Dev hidream	width, height, num_images
hidream-i1-full	Full hidream	width, height, num_images
ai-anime-generator	Anime Generator tools	width, height
wan2.1-text-to-image	Text to Image wan2.1	width, height
flux-kontext-pro-t2i	Kontext Pro T2I kontext	aspect_ratio
flux-kontext-max-t2i	Kontext Max T2I kontext	aspect_ratio
gpt4o-text-to-image	Text to Image gpt	aspect_ratio, num_images
midjourney-v7-text-to-image	Text to Image midjourney	speed, aspect_ratio, variety, stylization
flux-schnell	Schnell flux	width, height, num_images
gpt-image-2-text-to-image	GPT Image 2 (Text to Image) openai	aspect_ratio, resolution, quality
bytedance-seedream-v3	Text to Image v3 seedream	aspect_ratio
qwen-image	Text to Image qwen	aspect_ratio, num_images
ideogram-v3-t2i	v3 Text to Image ideogram	render_speed, style, aspect_ratio, num_images
nano-banana	Text to Image nano	aspect_ratio
google-imagen4	Imagen 4 google	aspect_ratio, num_images
google-imagen4-fast	Imagen 4 Fast google	aspect_ratio, num_images
google-imagen4-ultra	Imagen 4 Ultra google	aspect_ratio
sdxl-image	Text to Image sdxl	width, height
bytedance-seedream-v4	Text to Image v4 seedream	aspect_ratio, resolution, num_images
hunyuan-image-2.1	Text to Image v2.1 hunyuan	width, height
chroma-image	Text to Image chroma	width, height
flux-krea-dev	Krea Dev flux	aspect_ratio, num_images
perfect-pony-xl	Text to Image pony	width, height
neta-lumina	Text to Image neta	width, height
wan2.5-text-to-image	Text to Image wan2.5	width, height
hunyuan-image-3.0	Text to Image v3.0 hunyuan	width, height
leonardoai-phoenix-1.0	Phoenix 1.0 T2I leonardoai	aspect_ratio
leonardoai-lucid-origin	Lucid Origin T2I leonardoai	aspect_ratio
reve-text-to-image	Text to Image reve	aspect_ratio
grok-imagine-text-to-image	Text to Image grok	aspect_ratio
nano-banana-pro	Text to Image Pro nano	aspect_ratio, resolution
kling-o1-text-to-image	Text to Image [Pro] kling-o1	aspect_ratio, resolution, num_images
z-image-turbo	Text to Image Turbo z-image	width, height
flux-2-dev	Text to Image [Dev] flux-2	width, height
flux-2-flex	Text to Image [Flex] flux-2	aspect_ratio, resolution
flux-2-pro	Text to Image [Pro] flux-2	aspect_ratio, resolution
vidu-q2-text-to-image	Text to Image vidu-q2	aspect_ratio, resolution
bytedance-seedream-v4.5	Text to Image seedream-v45	aspect_ratio, quality
gpt-image-1.5	Text to Image gpt-1.5	aspect_ratio, quality
wan2.6-text-to-image	Text to Image wan2.6	width, height
flux-2-klein-4b	Text to Image [Klein 4B] flux-2	aspect_ratio
flux-2-klein-9b	Text to Image [Klein 9B] flux-2	aspect_ratio
z-image-base	Text to Image Base z-image	aspect_ratio, strength
seedream-5.0	Seedream 5.0 bytedance	prompt only

Training4 models

Model slug	Variant	Extra parameters
flux-dev-lora	Dev LoRA flux	model_id, width, height, num_images
wan2.1-lora-i2v	Image to Video (LoRA) wan2.1	lora_list, aspect_ratio, resolution, quality
wan2.1-lora-t2v	Text to Video (LoRA) wan2.1	lora_list, aspect_ratio, resolution, quality
sdxl-lora	LoRA sdxl	lora_list, width, height

Audio to Video11 models

Model slug	Variant	Extra parameters
sync-lipsync	sync lipsync	prompt only
latent-sync	latent lipsync	prompt only
creatify-lipsync	Creatify lipsync	prompt only
veed-lipsync	Veed lipsync	prompt only
wan2.2-speech-to-video	Audio to Video wan2.2	resolution
infinitetalk-image-to-video	Image to Video infinite-talk	resolution
kling-v1-avatar-standard	Standard A2V kling-v1	prompt only
kling-v1-avatar-pro	Pro A2V kling-v1	prompt only
kling-v2-avatar-standard	Standard A2V kling-v2	prompt only
kling-v2-avatar-pro	Pro A2V kling-v2	prompt only
ltx-2-19b-lipsync	Audio to Video ltx	resolution

Text to Text8 models

Model slug	Variant	Extra parameters
gpt-5-nano	GPT5 Nano Text gpt	prompt only
sora2-storyboard	Storyboard sora2	duration
gpt-5-mini	GPT5 Mini Text gpt	prompt only
text-passthrough	Text to Text text	make_input
any-llm	Text to Text llm	system_prompt, model, reasoning, priority
openrouter-vision	Image to Text llm	images_list, system_prompt, model, reasoning
agentic-architect	Architect workflow	prompt only
agent-chat	Agent Chat infra	message, conversation_id

Image to 3D9 models

Model slug	Variant	Extra parameters
meshy-6-image-to-3d	Image to 3D meshy	should_texture, topology, target_polycount, should_remesh
meshy-6-multi-image-to-3d	Image to 3D meshy	images_list, should_texture, topology, target_polycount
meshy-6-text-to-3d	Image to 3D meshy	mode, topology, target_polycount, should_remesh
tripo3d-h31-image-to-3d	Image to 3D tripo3d	texture, texture_quality, geometry_quality, pbr
tripo3d-h31-multiview-to-3d	Image to 3D tripo3d	images_list, texture, texture_quality, geometry_quality
tripo3d-h31-text-to-3d	Image to 3D tripo3d	texture, texture_quality, geometry_quality, pbr
tripo3d-p1-image-to-3d	Image to 3D tripo3d	texture, face_limit
tripo3d-p1-text-to-3d	Image to 3D tripo3d	texture, face_limit
meshy-image-to-3d	Image to 3D (Meshy direct) meshy	prompt only

How it compares

Why Loometo, and not the others.

	Loometo	ComfyUI	Vibe-Workflow	Runway
Visual node canvas	✓	✓	✓	✕
Your coding agent builds it (MCP)	✓	✕	✕	✕
Audio, voice clone & lip sync	✓	partial	✕	partial
Image → 3D with STL / GLB export	✓	✕	✕	✕
Model-agnostic (276, one key)	✓	local weights	a few	own models
Bring your own keys, pay at cost	✓	free / local	—	subscription
Runs 100% local, no telemetry	✓	✓	—	✕ cloud
Ready-made templates	22	✕	a few	presets
In-app cost labels before you run	✓	n/a	✕	credits

Positioning, not a takedown: ComfyUI runs local model weights; Vibe-Workflow is a lighter node builder; Runway is a closed model lab. Loometo is the open orchestration layer over everyone's models.

License & branding

Open code. Protected brand. The Firefox model.

The same approach Firefox and Chromium use: take the code and do anything with it, just don't ship it as "Loometo."

Code APACHE-2.0

Use, copy, modify, redistribute
Commercial use: sell it, build a business on it
Fork it, run it for clients, ship products on it
Say it's "built on Loometo", factual reference is fine

Brand TRADEMARK · RESERVED

Don't name your fork "Loometo" or a lookalike
Don't reuse the Loometo logo or brand as your own
Don't imply your version is official or endorsed
Keep the NOTICE file: credit to Dinimiciuil Labs travels with the code