OMNIVAIOMNIVAI

Model Library

Every frontier model. One studio.

29 selectable models for video, image, audio and upscaling — official integrations, real outputs, no API plumbing. Every name and price below is read from the live model registry, so what you see here is exactly what opens in the studio.

Video Models

Cinematic motion. Frame by frame.

Seedance 2.5, Veo 3.1, Kling O3 and V3, Wan 2.6 and Hailuo 02 — text-to-video, image-to-video, reference video, motion control, lip-sync and native audio.

Video · up to 30s

Seedance 2.5

Premium · up to 30s · edit and extend existing video

782cr · Try it
Video · native audio

Veo 3.1 Pro (Stable)

Highest quality stable variant

995cr · Try it
Video · motion

Kling O3 Pro

All O3 modes at up to 4K — production grade

352cr · Try it
Video · fast pass

Wan 2.6

Text-to-video · open-source flagship

209cr · Try it

Frames rendered in OmniVAI to illustrate the category — not a benchmark of the individual model.

All 17 in this category

Seedance 2.0 StandardT2V · I2V · Reference — up to 4K, native audio1323crSeedance 2.5Premium · up to 30s · edit and extend existing video782crGemini Omni Flash VideoLow-cost bulk video · 4 modes232crKling V3 StandardT2V · I2V · Standard mode — realistic motion265crKling V3 ProT2V · I2V · Professional mode — hero-grade realism265crKling V3 Motion ControlMotion transfer — drive a character image with a reference video400crPixverse Lipsync v4.5Mouth-perfect lip-sync — drive any video with any audio160crKling Lipsync v2High-fidelity lip-sync for close-ups and dialogue334crKling O3 StandardT2V · I2V · Reference · Video Edit · Custom Element — 1080p352crKling O3 ProAll O3 modes at up to 4K — production grade352crVeo 3.1 Fast (Stable)Per-second · best price/quality374crVeo 3.1 Pro (Stable)Highest quality stable variant995crWan 2.6Text-to-video · open-source flagship209crWan 2.6 Image-to-VideoAnimate any image209crWan 2.6 Reference-to-VideoMulti-reference · identity locked209crHailuo 02 StandardText-to-video · balanced quality185crHailuo 02 DirectorCamera-direction tags · cinematic control317cr

Image Models

Stills that ship. Pixel-perfect.

Seedream 5.0, Nano Banana Pro and 2, Midjourney V8.1 and V7, GPT-Image-2 High — photoreal people and products, style reference, legible type and character consistency.

Image · portrait example rendered in OmniVAI
Image · portrait

Nano Banana Pro

3D · stylized · products

42cr · Try it
Image · product example rendered in OmniVAI
Image · product

Nano Banana 2

Gemini 3.1 Flash Image · 0.5K→4K · search-grounded

32cr · Try it
Image · scene example rendered in OmniVAI
Image · scene

Seedream 5.0

Photoreal · people · ads

23cr · Try it
Image · editorial example rendered in OmniVAI
Image · editorial

Midjourney V7

Generate · 4 images per run

27cr · Try it

Frames rendered in OmniVAI to illustrate the category — not a benchmark of the individual model.

All 6 in this category

Seedream 5.0Photoreal · people · ads23crNano Banana Pro3D · stylized · products42crNano Banana 2Gemini 3.1 Flash Image · 0.5K→4K · search-grounded32crMidjourney V8.1Highest-craft stylization · 4 images per run27crGPT-Image-2 HighOpenAI · hero quality60crMidjourney V7Generate · 4 images per run27cr

Audio Models

Sound that scores. Voice & music.

Suno V5 writes full tracks with vocals. ElevenLabs v3 covers multilingual narration, instant voice clones and sound effects.

All 4 in this category

ElevenLabs v3 — Multilingual TTSStudio-grade text-to-speech · 32 languages · emotion tags4crElevenLabs v3 — Instant Voice CloneClone any voice from a 30-second sample4crElevenLabs v3 — Sound EffectsGenerate sound effects from a short text prompt14crSuno V5 — Music & vocalsFlagship music model · songs & instrumentals20cr

Upscale

Same shot. More resolution.

Topaz Image and Topaz Video recover detail and push a finished frame or clip up to 4K — no re-roll, no lost identity.

All 2 in this category

Topaz Image UpscaleIndustry-standard 4K image upscaler27crTopaz Video UpscaleStudio-grade 4K video upscaler54cr

Where they run

You don't pick a model. You pick a studio.

Every studio follows the same three stages — Brief → Build → Final — and Omni routes each stage to the right model above.

Cinematic StudioSeedance · Kling · Veo · WanUGC Studiotalking creators · lip-syncProduct Ads Studioproduct hero · TVCBrand Studiobrand sheet · kitsPhoto StudioSeedream · Midjourney · Nano BananaSocial Studiovertical cutdownsMusic StudioSuno V5 · score & vocalsCharacter Labconsistent characters

One subscription. Every model.

Skip the per-API juggling. Open the studio and pick anything in this library — credits roll across models.

Open Studio →