Krea 2 Open-Source · Brand Campaign

Atelier Nordtrain your brand aesthetic

One sentence of art direction becomes a cohesive eight-shot campaign — then a packshot, a motion teaser, and a finished reel. Built on Krea 2 Open-Source, the most aesthetic open image model on the network, and the first open base you can train.

Base model
krea-2-os
Campaign render
8 shots · ~3s each
Cost
~$0.25 e2e
Finishing
bg-remove · i2v · ffmpeg
Vertical
Fashion (generalizes)
krea-2-oskrea-2-lorakrea-2-lora-training bg-removekling-v3-turbo-i2vffmpeg-concatffmpeg-export
The pitch. A bare text-to-image model gives you eight pretty-but-unrelated pictures. Krea 2's out-of-the-box taste — light, palette, grain, restraint — makes eight shots read as one campaign from a single locked aesthetic sentence. Lock the look further into a brand LoRA and every future shot stays on-brand by construction. This is the open-weights image model that finally looks art-directed before you touch it.

The hero krea-2-os · one locked aesthetic, eight shots

Every shot shares one byte-identical art-direction prefix — “ivory, sand, warm stone & oat palette; soft diffused north-window light; sculptural wool & linen; quiet luxury; fine film grain; Vogue-cover restraint; medium-format.” Only the trailing subject changes. The palette, light and grain hold across all eight.

ATELIER NORD hero portrait — model in an ivory wool coat by a window
full look
02 · full look
detail texture
03 · detail / texture
lifestyle
04 · lifestyle
accessory still life
05 · accessory
atmospheric wide
06 · atmospheric
back look at window
07 · back look
beauty close-up
08 · beauty
packshot — scarf and gloves on transparent background
→ bg-remove packshot

All eight rendered on sdk.daydream.monster via krea-2-os (fal-ai/krea-2/turbo) — ~3 seconds per shot, ~$0.25 for the set.

Finish it the same campaign, made deliverable

motion teaser · kling-v3-turbo-i2v from the hero still (5s)
campaign reel · ffmpeg-concat (8 stills + teaser) → 4:5, ~18s

The packshot above is a one-call bg-remove on the accessory still — a clean transparent cutout for e-commerce. The teaser animates the hero portrait; the reel stitches the campaign into a platform cut. Add sonilo-t2m for a licensed soundtrack and ffmpeg-export to fan out 9:16 / 1:1 / 16:9.

Lock the look krea-2 LoRA · the brand-consistency unlock

Eight shots from one sentence already feel like a campaign. To make it permanent — so every future look is on-brand by construction across 8, 80, or 800 shots — freeze the aesthetic into a LoRA. Krea 2 Open-Source is the first open base on the network you can train, and the one to pick when look matters more than a photoreal identity.

01 · SEED
Render the look
8–16 cohesive shots from one locked aesthetic sentence (krea-2-os). The set above.
02 · TRAIN
Freeze it to a LoRA
submit_lora_train · base_model:"krea-2" → trains on fal-ai/krea-2-trainer. ~6–8 min, ~$22, once.
03 · SHIP
Generate on-brand forever
create_media · model_override:"krea-2-lora" — every new look inherits the house aesthetic.
# Train your brand aesthetic into a reusable LoRA — across every surface
# MCP
submit_lora_train({ zip_url, trigger_word:"ATELIERNORD", base_model:"krea-2", steps:500 })
apply_lora({ weights_url, trigger:"ATELIERNORD", prompt:"a new SS26 look", model:"krea-2-lora" })

# CLI
livepeer lora train --zip <url> --trigger ATELIERNORD --base krea-2 --steps 500
livepeer lora apply <weights_url> --trigger ATELIERNORD --prompt "a new SS26 look" --model krea-2-lora

# Webapp: right-click canvas → Train LoRA → base = "Krea 2 (Open-Source)" → /lora apply <name>

Honest status (2026-06-24): the krea-2-lora-training cap is registered, accepts jobs and meters them; full weight-retrieval through the SDK currently hits a known 202-polling follow-up (the orchestrator returns HTTP 202 for async training and the SDK's status poller doesn't yet treat 202 as “still running”). The recipe above is wired across MCP + CLI + webapp and lights up the moment that SDK fix lands. The campaign and finishing passes shown on this page are fully proven, end-to-end.

It generalizes same pipeline, new brief

The aesthetic sentence is the whole product. Swap it and the same pipeline ships a different vertical:

REAL ESTATE
Listing campaign
“warm golden-hour interiors, wide-angle, natural oak & limestone, lifestyle staging.” Shots become rooms; the teaser becomes a walk-through.
TRAVEL
Destination set
“sun-drenched film-stock postcards, Kodak Portra, golden coast light.” Shots become destinations; LoRA locks a publisher's signature grade.
EDUCATION
Course visuals
“clean isometric explainer plates, soft pastel, generous whitespace.” Shots become concept panels; one LoRA keeps a whole curriculum on-brand.

Why it matters. Aesthetic used to be the moat closed apps charged for. Krea 2 Open-Source puts the most tasteful open image base on a metered network — and makes it trainable, so a small brand can lock its look once and out-produce a studio. The campaign on this page cost about a quarter and ran in under a minute.

Follow it yourself. Open the Train Your Brand Aesthetic playbook, fill in one aesthetic sentence, and run it.