Where 1.0 fell short
Vephon Studio 1.0 was the right idea — persona consistency in an AI content tool — but in practice it didn't fully deliver on the promise. You'd create a persona, generate a few posts, and within a week you were back to fighting the same drift problems every other AI tool has. The character would shift. The brand voice would drift. The formats wouldn't match each platform. You'd end up doing the same manual work you were trying to avoid.
We spent the last quarter listening to users and ripping out everything that wasn't carrying its weight. What we shipped in 2.0 is not a feature update. It's a rewrite organised around six pillars.
The six pillars
- The Cast — multi-angle character refs so a persona is a real digital cast member, not a face that drifts.
- The Edit Bay — Smart Edit (FLUX Kontext) for mask-free in-place edits so iteration stops being slot-machine roulette.
- The Score — Eleven Music and Eleven v3 expressive voice so AI video stops being silent and AI narration stops being flat.
- The Channel — Multi-Platform Variants so one prompt produces IG, Reels, TikTok, X, LinkedIn, and standalone audio in one batch.
- The Director — Live Mode (coming in 3.0) for voice-driven creation with Gemini Live.
- The Atelier — the admin surface our internal team needs to run the business: feature flags, moderation queue, AI cost board, audit log, user management.
Each pillar replaces a piece of busy-work that was the actual reason "AI for marketing" felt frustrating in practice.
The wedge: persistence
The thread that ties the pillars together is persistence.
Every other AI tool starts from scratch every session. Same prompt, different output. Same persona, different face. Same brand, different voice. Vephon's wedge is that the cast member you build, the brand kit you define, the references you upload, and the moderation verdicts admins record — all of it persists and quietly influences every generation forever.
That's the difference between an AI image generator and a creative studio you can build a brand on.
What we didn't ship
We deliberately punted on a few things:
- Real-time collaboration. Multiplayer editing is expensive and the pillars above are the wedge. We'll add it when the studio has organic demand for it, not before.
- A marketplace UI. The database tables are there, but a real marketplace is its own product. We ship it in 2.1 if the demand is there.
- Every social platform OAuth. Instagram, X, and LinkedIn ship properly. TikTok and YouTube come in 2.1.
Saying no to those let us ship the six pillars properly instead of six half-built features.
What's next
We're heads-down on Live Mode (voice-driven director with Gemini Live), workspace-level project switching, and Kling 3.0's reference video lip-sync. Expect the next major release at the end of summer.
In the meantime, if you're building a brand or a digital persona that needs to live across IG, Reels, LinkedIn, and X without drift — Vephon 2.0 is the tool we wanted to build the first time. Try it, build a real campaign with it, and tell us where it still hurts. The roadmap responds to what users tell us, not what we guess.
— The Vephon team