Black Forest Labs
Generate a quarter of brand-aligned visuals on Martini using FLUX.2 — campaign hero, social tiles, blog headers, and deck imagery from one anchor canvas. FLUX.2 is the prompt-fidelity pick: it renders specific brand props, palettes, and compositions almost literally, without the embellishment Midjourney layers on. Pair it with a saved brand prompt block, fan that into every downstream node, and the model holds palette, lighting language, and compositional intent across hundreds of generations.
FLUX.2 reads natural language, not keyword soup. Write a brand prompt block once: "Editorial photography style, soft window light from camera left, warm beige and deep moss palette, shallow depth of field, magazine-quality crop with generous negative space." Save it in a Text node on the canvas; every FLUX.2 node concatenates it with the per-asset subject prompt.
FLUX.2 has a Pro tier with tighter prompt adherence and higher detail rendering, plus a base tier for default work. Pin Pro on the campaign hero and any large-format placement; pin base on lifestyle social tiles and exploratory drafts. The cost difference is measurable across a quarter — set the tier per asset class once.
On the canvas, wire the brand Text node into multiple FLUX.2 nodes — campaign hero (3:4), social square (1:1), Reels vertical (9:16), blog header (16:9), deck cover (16:9). Each FLUX.2 node concatenates the brand block with its own subject prompt. Brand voice is identical across the five outputs because the prompt prefix is identical.
FLUX.2 is sensitive to prompt changes — even swapping one adjective shifts composition. Use this: generate a baseline, then change one variable per generation (light direction, palette accent, prop count). Keep the brand block untouched; only edit the per-asset subject prompt. Place multiple FLUX.2 nodes side by side on the canvas to compare results.
Every 10-20 generations, drop the latest FLUX.2 output next to the original brand reference (or the first locked asset). Check palette, light direction, and prop styling. If drift is visible, re-read the brand prompt block — drift in FLUX.2 is almost always a prompt issue, not a model issue. Tighten the language and re-run.
Once a campaign ships, save the canvas — brand Text node, fan-out FLUX.2 nodes, parameter pins. Next quarter, swap the subject prompts and re-run. The brand pipeline becomes reusable: same canvas, different campaign, same aesthetic. This is what compounds across multiple quarters of brand work.
Standing brand prompt prefix. Wire this Text node into every FLUX.2 node on the canvas — the brand voice flows through one source of truth.
Editorial photography style, soft window light from camera left, warm beige and deep moss palette, shallow depth of field, magazine-quality crop with generous negative space, single ceramic vase center frame on a beige linen surface, 3:4 aspect
Social square that inherits the prefix. The subject extends without overriding the brand block.
[brand prefix] + Lifestyle morning scene, hands lifting a ceramic mug from a wooden tray, soft golden window light, 1:1 social aspect
Blog header in the same brand world. Prefix locks studio voice; subject prompt extends scope.
[brand prefix] + Wide composition of a ceramic studio interior, multiple vases drying on a wooden shelf, late afternoon raking light, atmospheric depth, 16:9 aspect
Deck cover with copy room baked in. Pin FLUX.2 base tier — Pro is overkill for low-detail covers.
[brand prefix] + Minimalist single vase silhouette against pale gradient, generous negative space top-right for headline copy, editorial restraint, 16:9 aspect
FLUX.2 has no style presets — the prompt is your only control surface. Invest time in the saved brand block; reward compounds across the campaign.
Pin Pro tier for hero work, base for everything else. Switching mid-campaign breaks consistency without obvious benefit.
Write spatial language explicitly ("left half of the frame," "single vase center frame"). FLUX.2 follows literal positions more reliably than any other image model in Martini.
Avoid in-image text — chain to Ideogram for headlines. FLUX.2 can render typography but Ideogram wins on legibility for production work.
Iterate by changing one element per generation, never the whole prompt. FLUX.2 will shift composition on every word change.
Match aspect ratio to the placement (3:4 hero, 1:1 social, 9:16 vertical, 16:9 deck). FLUX.2 composes differently per ratio.
FLUX.2 returns 1024-2048 wide outputs depending on tier. Prompt-fidelity outputs hold composition, palette, and prop placement at ~90% consistency across 50+ generations from the same brand block. Generation time: base 15-30s, Pro 30-60s. Output drops onto the canvas as a single selected frame. Route the chosen output into ad-creative, social-graphic, or video starting-frame nodes downstream — the same brand block ensures voice continuity across modalities.
Connect FLUX.2 with other AI models on Martini's infinite canvas. No GPU required — start free.
Get Started FreeMidjourney
Ship a quarter of brand-aligned visuals — campaign hero, social tiles, blog headers, deck imagery — in one Martini canvas using Midjourney v7. The model is the strongest pick for editorial-photography aesthetics and painterly brand worlds where mood and atmosphere carry the work. Lock a single style reference image, fan that anchor into every downstream node, and Midjourney holds palette, lighting, and tonal voice across hundreds of generations without redrafting the prompt each time.
View guideBuild a quarter of brand-aligned imagery — hero campaign frames, social tiles, blog headers, presentation imagery — on Martini using Imagen 4. Imagen 4 wins on photographic realism and on prompt adherence: it renders specific brand props, lighting setups, and color palettes with near-literal accuracy. Pair it with a saved style prompt block and a brand color reference, fan that into every downstream node, and the same canvas ships consistent campaign visuals across hundreds of generations.
View guide