[/services/content] · AI content & creatives
An AI pipeline for product photos, video for Reels / TikTok / Meta Ads, and voiceover in Ukrainian and 20+ languages. We adapt it to the specs of each ad platform. Testing hypotheses no longer depends on a photographer's schedule.
online
Valeriy Yosypovych
AI manager · 45 y.o.
What needs generating — product photos, Reels video, voice? I'll give you timelines and pricing.
// one conversation — in the hero and in the bubble bottom-right
[01] a filter before we start
[02] 4 types of AI content
Most clients start with 1-2 types, then add the rest.
AI product photos based on your own imagery. Not "generate something" from scratch — we train a model on your real product photos, then generate new angles, models, and scenes without a studio. We can "try" the product on different people, different backgrounds, and different conditions (day/night, indoor/outdoor).
Tools: Nano Banana (Google), Gemini Pro Vision, GPT-image, ComfyUI with custom models
$40 per creative à la carte. Retainer from $2,400 / mo (~100 creatives).
AI-generated 5-15-second clips from a brief. Cinematic motion, text overlays, music. Not "cringe 2023 AI video" — current quality (Sora, Veo, Runway). We work in three formats: 9:16 (Reels / TikTok / Stories), 1:1 (Meta Feed), 16:9 (YouTube).
Tools: Sora, Google Veo 3, Runway Gen-3, Kling AI, Suno (for sound)
$160-240 per 5-15 sec video. Retainer bundle (50 videos / mo) — from $5,000.
Voiceover in Ukrainian, English, Russian, Polish, German + 20 more languages. We can clone your manager's / narrator's voice from a 30-sec recording and voice all your clips with it. Tone is adjustable (energetic / calm / professional / friendly).
Tools: ElevenLabs, OpenAI TTS, Whisper (for transcripts)
from $100 per 60 sec of voiceover. Voice cloning is a one-time $400.
One concept → 20 versions across platforms and targeting. The same product, but different taglines (for cold, warm, hot audiences), different visuals (student, corporate, mom-on-leave), different formats (9:16 for Reels, 1:1 for Meta Feed, 1080×1920 for Stories). An industrial performance machine.
Tools: A combination of everything above + our orchestration via Trinity Runtime
Retainer only, from $3,600 / mo. We tune it to your performance structure.
[03] pricing
A retainer works best. À la carte — for a test or a one-off task.
| Type | Price (from) | Timeline |
|---|---|---|
| AI product photo (à la carte) | $40 | 1-2 days |
| AI video 5-15 sec (à la carte) | $160-240 | 2-3 days |
| Voiceover 60 sec | $100 | 1 day |
| Voice cloning (one-time) | $400 | 2-3 days |
| Pipeline setup for your brand (one-time) | $600-1,000 | 1 week |
| Photo-pipeline retainer (~100 creatives / mo) | from $2,400 / mo | starts within a week |
| Video-bundle retainer (50 videos / mo) | from $5,000 / mo | starts within a week |
| Adaptive performance pipeline (retainer only) | from $3,600 / mo | 1-2 weeks of setup |
| Free 30-min audit | $0 | after booking |
* Setting up the pipeline for your brand (brand book, voice, typical scenes) is a one-time $600-1,000. AI model costs (Sora, ElevenLabs) are included in the creative price.
[04] in your niche
Find your industry — you'll see a typical setup.
Product creatives on a conveyor: 30-50 new visuals a week across all products. Fast A/B testing of new angles and new audiences (young moms / professionals / students). Every format: feed, story, reel, banner.
Testing pace up 3-5x
An AI content pipeline for your own clients. Replaces a motion designer + photographer + copywriter for the hypothesis-testing part. Runs in the background. Partnership details on /for-agencies.
White-label from $3,600 / mo
Social creatives: before / after (based on your real photos + AI styling to preserve anonymity), team stories, video walkthroughs of procedures. No live photoshoots every week.
5-10 ready creatives a week
Creatives for launch campaigns: 50-100 variants for cold / warm / hot audiences. Voiceover of tutorials and the curriculum in your own (cloned) voice. Content marathons on TikTok.
CPL down 2-3x through fast testing
AI visualization of properties: apartments, houses, landscaping. Interiors in different styles (loft, scandi, classic) — the client picks their own. Exteriors across seasons and lighting. No rendering by a 3D studio.
An alternative to rendering: $40 instead of $1,000
[05] we live on what we sell
A performance agency with 5+ years of practice. Each client project means 20-40 creative hypotheses a month. Before the AI pipeline we'd have hired 2 motion designers and a photographer. Now the same volume is handled by one pipeline + one creative director on top.
100+
creatives a month (our steady volume)
5×
faster than traditional production
−70%
cost per creative vs. a full-cycle agency
* figures from real MaxICo Agency campaigns. Concrete examples are in the cases block below.
[06] cases
[07] common objections
It depends on the type. Product photos on a neutral background — indistinguishable. Photos with people (especially close-up faces) — an experienced viewer will notice AI artifacts. So we use AI where it's strong (products, interiors, abstract concepts) and avoid it where it's weak (face close-ups, complex hand anatomy). Facebook / TikTok / Google don't require labeling AI content in ads (as of 2025), but some countries (the EU) may introduce disclosure rules in the future.
We don't try to replace a photographer where they're better: brand campaigns with people, reportage, documentary shoots, high fashion. We work on performance tasks: product creatives for testing, Reels video at scale, voice for mass voiceover. The best mix: one quality photoshoot a year + an AI pipeline for weekly tests.
The standard terms of AI models (Sora, ElevenLabs, Nano Banana) allow commercial use of generated content if you're on a paid plan (we pay for it). The generated photo is yours — you can use it in ads, on your site, in your product. Exceptions: you can't generate the likeness of real people without their permission, and you can't copy the protected styles of other brands.
Yes. Meta, TikTok, and Google allow AI-generated content in ads. The restrictions are standard (as for any creative): no nudity, no weapons without context, no deception, no unverified medical claims. AI generation in itself is not blocked.
Standard cases: a photo — 1-2 days from the brief. A 5-15 sec video — 2-3 days. Voice — 1 day. An adaptive pipeline (one concept in 20 versions) — 3-5 days. On a retainer the client already has templates and a tuned pipeline, so individual creatives come out within hours.
Midjourney + manual editing is one-off craft. One creative = one hour of a designer's work. It doesn't scale, it isn't repeatable. We build a pipeline: a brand model trained on your products, scripts that automate the routine (background removal, ratio cropping, text overlay), orchestration via Trinity Runtime. One concept automatically becomes 20 versions across platforms. It's engineering, not artistry.
[08] let's talk
On the audit: we take 1-2 of your products / services and generate 3-4 creative variants right there on the call. You'll see the quality and potential with your own eyes. No slide decks.
// if AI won't cover your tasks — on the call we'll honestly tell you where a live photographer / motion designer is better.
