Montage
From web/mobile app to social posts: Navin Montage (demo, Veo, Lyria)
August 9, 2026 · 3 min read · Navin team
Build web and mobile in Navin, then let AI create project-linked videos and images (Nano Banana, Veo 3.1 Fast, Lyria, Qwen STT) for YouTube Shorts Reels TikTok - no auto-publish.
Navin's strength is not “another video tool”. It is the full loop: you build the web and mobile app in one workbench (Code, Preview, Mobile), then the AI creates and montages videos and images tied to that project for social networks - same workspace, no invented product.
The Montage studio (#/montage, /montage) runs that second half of GTM: real UI footage first, AI creatives next, exports for YouTube / Shorts / Reels / Feed / TikTok / LinkedIn, no auto-publish.
Docs: Montage · Media models · Toolchain
From app to posts: the loop
- Build - web (Vision 360 / Preview) and/or mobile (
/mobile, Android preview). - Link the project in Studio → Montage.
- Film the real happy path (you or the agent with URL + steps) →
marketing/montage/demos/. - Package social ratios (FFmpeg) →
marketing/montage/exports/. - Enrich if needed: product stills, AI B-roll, Lyria music, TTS voiceover, STT captions.
- Review, then post manually - or move to
/adsonly with explicit approval.
Source of truth stays your product. The AI speeds the social kit around the linked repo.
Live demo vs AI creatives
| Source | When to use it |
|---|---|
Browser capture (record_start / record_stop) | Authentic web/mobile happy-path showcase |
| AI image (default Nano Banana 2) | Packshots, lifestyle, covers tied to the project brief |
| AI video (default Veo 3.1 Fast) | B-roll, transitions, teaser when UI is not filmable yet |
| Music (default Lyria Clip 30 s) | Cheap bed before a Pro track |
| TTS / STT | Voiceover and captions on the demo |
Brief agent requires URL + steps - not a vague prompt. You can also record yourself in the Agent browser.
Models in use (managed Plus+ catalog)
Defaults shipped with the Navin media stack (pickable in Montage → AI media providers):
| Modality | Default | Curated alternatives |
|---|---|---|
| Image | google/gemini-3.1-flash-image (Nano Banana 2) | Nano Banana Pro, Seedream 4.5 (~$0.04/image) |
| Video | google/veo-3.1-fast | Veo 3.1, Seedance 2.5, Kling v3.0 Pro, Seedance 2.0 Fast |
| Music | google/lyria-3-clip-preview ($0.04 / 30 s clip) | Lyria 3 Pro ($0.08 / song) |
| TTS | google/gemini-3.1-flash-tts-preview | Fish Audio, Grok Voice TTS |
| STT | qwen/qwen3-asr-flash-2026-02-10 | GPT Transcribe, Parakeet, MAI-Transcribe, Deepgram Nova-3 |
FFmpeg packaging of an already recorded demo can run with zero media API spend. Managed video/music batches need Plus+ or BYOK. Full slugs and pricing notes: media models.
Platform exports
Built-in profiles: YouTube Landscape 1920×1080, Shorts / Reels / TikTok 1080×1920, Feed 1080×1080, LinkedIn 1920×1080. profiles=all adds YouTube 4K and cinematic 21:9.
Files under the project: marketing/montage/demos/, exports/, captures/, compositions/ + montage-report-*.html.
Montage vs Marketing vs Ads
| Studio | Role in the loop |
|---|---|
| Code / Mobile | Build the web and mobile app |
| Montage | Project-linked video proof + images + social exports |
Marketing (/campaign) | Copy, posts, messaging campaigns |
Ads (/ads) | Paid MCP delivery - only after approval |
FAQ
Can AI replace a real UI demo?
Not for a serious product showcase. Prefer live capture; use Veo / Seedance / Kling as complement.
Do I need HyperFrames to post on TikTok?
No. Social crops = demo + FFmpeg. HyperFrames is HTML→MP4 composition.
Is Free enough?
Toolchain (FFmpeg, package an existing demo): yes. Managed media generations: Plus+ or BYOK.
Download · Montage docs · Models · Install FFmpeg / HyperFrames · Code module · Mobile
Try Navin on your machine
Local agent, cross-platform. Code, debug, scrape, leads, security and review - without leaving Navin.