What's New

Vephon Studio 2.0 — the Studio Update

This is the biggest release we've ever shipped. We re-thought the product from the ground up around six pillars that turn Vephon from "an AI image generator with personas" into a full creative studio:

The Cast — multi-angle character refs

Your personas are no longer just a face and a body. You can now attach multi-angle references — front, three-quarter, side, back, close-up, full body — with their own expressions (neutral, smile, serious, laugh, focus, …) and outfits. The image model picks the best matching set per generation so your digital cast members stay recognisable across every shot.

Find it under any persona → Character refs.

The Edit Bay — Smart Edit (FLUX Kontext)

Open any generation in the gallery and hit Smart Edit. Plain English: "remove the cup", "change her outfit to red", "make the sky golden hour". Powered by Black Forest Labs' FLUX Kontext, the rest of the composition is preserved — faces, lighting, layout — so you can iterate without re-rolling the whole frame.

The Score — Eleven Music + Eleven v3

• Music generation: request a track by style ("upbeat lofi for a coffee shop, 110 BPM, vinyl warmth"), with optional lyrics that can use a cloned voice.
• Expressive narration: Eleven v3 with emotion tags ([whispering], [excited], [sighs], …) brings performance direction into TTS.

The Channel — Multi-Platform Variants

One prompt → IG square, IG Reel, IG carousel, X text, X image, LinkedIn text, LinkedIn image, standalone voiceover — all spawned as independent content jobs that land in the gallery as a single group. New top-level nav entry: Multi-Platform.

The Atelier — Brand Hub

Brand kits are back, and they actually apply. Define colors, fonts, brand voice, tone words, do-say / don't-say phrases, image style prompt, forbidden subjects, voice clone id — and Vephon weaves them into every generation silently. Mark one as default and every job picks it up.

Video 3.1 + scene extension

Video generation moves to Veo 3.1, with three tiers: Light ($0.05/sec), Fast ($0.15/sec), and flagship 4K with native audio ($0.40/sec). A new Extend action takes an existing clip and adds 8–30s anchored to the last frame so the new footage continues the same scene.

Trust & Safety

A new Moderation Queue under admin captures provider safety flags, user reports, and image-hash matches. Each event carries decision, severity, source, and categories — admins can allow / flag / block with notes, and decisions are immutable for audit.

Feature Flags

Server-evaluated flags with per-tier, per-user, and percentage-rollout gates. Toggle any feature without a redeploy from the admin console. Cache TTL is 15s so changes propagate within seconds.

Smaller wins

• Veo 3.1 pricing surfaces correctly in the credits ledger.
• Eleven v3 emotion-tag sanitizer drops unrecognized bracket text so scripts read naturally even with experimental markup.
• Pricing catalog now covers FLUX Kontext pro/max, Eleven Music, Veo 3.1 Light/Fast/Flagship, Grok-2 image, and Gemini 2.5 Flash thinking.
• Mobile app gains a Smart Edit screen and API clients for every new feature.

What's next (3.0)

We're heads-down on Live Mode (voice-driven director with Gemini Live) and a workspace-level project switcher with brand kit pinning. Expect both in the next release.

— The Vephon team

What's new in 1.0

After months of private testing, Vephon Studio 1.0 is now live on Google Play for Android. Here's what's available at launch:

Face-locked personas

Create a persona with a description. We generate a canonical reference image, and future outputs use that reference to preserve identity traits across content.

Reusable asset library

Create outfits, locations, and props once. Reference them with a #tag in any future prompt. The same asset appears consistently across all posts that use it.

Platform-native content generation

Generate complete posts, not just images. Instagram carousels with captions and hashtags. LinkedIn posts with professional tone. X tweets within character limits.

AI content orchestration

Describe what you want in plain language. Our AI planning layer breaks down your request, determines the optimal shot composition, writes the platform-appropriate caption, and selects relevant hashtags.

Voice generation

Design a voice for your persona from a text description. Generate voiceovers that match the character's personality. Voice cloning available on Creator plan and above.

Free tier

Free tier and full pricing details are available in-app on first launch.

---

What's coming in 1.1

• Multi-persona scenes (up to 5 characters in a single image)
• Video generation support (short clips)
• Improved 4K image quality
• Asset version history
• API access (Agency plan)

---

We're building in public. Share feedback at connect@vephon.com or in the community forum.