Skip to main content

​
Setup

import { createVarg } from "vargai/ai"

const varg = createVarg({ apiKey: process.env.VARG_API_KEY! })
All models are accessed through the varg client. 1 credit = 1 cent.
For detailed documentation, parameters, and examples for each model, see the Models section.

​
Quick Reference

CategoryRecommendedCodeCreditsDocs
Image (versatile)Nano Banana Provarg.imageModel("nano-banana-pro")5Details
Image (quality)Flux Provarg.imageModel("flux-pro")10Details
Image (editing)Nano Banana Editvarg.imageModel("nano-banana-pro/edit")5Details
Image (characters)Soulvarg.imageModel("soul")15Details
Image (cheapest)Grok Imaginevarg.imageModel("grok-imagine-image")3Details
Video (best)Kling V3varg.videoModel("kling-v3")150Details
Video (premium)Seedance 2varg.videoModel("seedance-2-preview")250Details
Video (budget)Wan 2.5varg.videoModel("wan-2.5")80Details
Video (cheapest)LTXvarg.videoModel("ltx-2-19b-distilled")50Details
Lipsync (simple)VEED Fabricvarg.videoModel("veed-fabric-1.0")100Details
Lipsync (best)Sync V2 Provarg.videoModel("sync-v2-pro")80Details
Speech (best)Eleven V3varg.speechModel("eleven_v3")25Details
Speech (multi-lang)Eleven Multilingual V2varg.speechModel("eleven_multilingual_v2")25Details
MusicMusic V1varg.musicModel()30Details
TranscriptionGroq Whisper Turbovarg.transcriptionModel("groq-whisper-large-v3-turbo")3Details
Image upscaleSeedVRvarg.imageModel("seedvr")10Details
Video upscaleSimavarg.videoModel("sima-video-upscaler")15Details

​
All models by category

​
Video Models

ModelCodeDurationCreditsDocs
Seedance 2varg.videoModel("seedance-2-preview")5, 10, or 15s250Details
Seedance 2 Fastvarg.videoModel("seedance-2-fast-preview")5, 10, or 15s150Details
Kling V3 Provarg.videoModel("kling-v3")3-15s (int)150Details
Kling V3 Standardvarg.videoModel("kling-v3-standard")3-15s (int)100Details
Sora 2varg.videoModel("sora-2")varies100Details
Sora 2 Provarg.videoModel("sora-2-pro")varies300Details
Sora 2 Remixvarg.videoModel("sora-2-remix")varies100Details
Grok Imaginevarg.videoModel("grok-imagine")1-15s100Details
Grok Imagine Editvarg.videoModel("grok-imagine-edit")varies100Details
Wan 2.5varg.videoModel("wan-2.5")3-10s80Details
Minimaxvarg.videoModel("minimax")5-10s80Details
LTXvarg.videoModel("ltx-2-19b-distilled")varies50Details

​
Image Models

ModelCodeCreditsDocs
Nano Banana Provarg.imageModel("nano-banana-pro")5Details
Nano Banana Editvarg.imageModel("nano-banana-pro/edit")5Details
Nano Banana 2varg.imageModel("nano-banana-2")5Details
Flux Provarg.imageModel("flux-pro")10Details
Flux Devvarg.imageModel("flux-dev")8Details
Flux Schnellvarg.imageModel("flux-schnell")5Details
Photavarg.imageModel("phota")9Details
Phota Editvarg.imageModel("phota/edit")9Details
Phota Enhancevarg.imageModel("phota/enhance")13Details
Grok Imagine Imagevarg.imageModel("grok-imagine-image")3Details
Recraft V3varg.imageModel("recraft-v3")10Details
Recraft V4 Provarg.imageModel("recraft-v4-pro")10Details
Qwen Anglesvarg.imageModel("qwen-angles")8Details
Seedream V4.5 Editvarg.imageModel("seedream-v4.5/edit")10Details
Soulvarg.imageModel("soul")15Details

​
Lipsync Models

ModelCodeCreditsDocs
VEED Fabric 1.0varg.videoModel("veed-fabric-1.0")100Details
Sync V2 Provarg.videoModel("sync-v2-pro")80Details
Sync V2varg.videoModel("sync-v2")50Details
OmniHuman V1.5varg.videoModel("omnihuman-v1.5")100Details

​
Upscale Models

ModelCodeCreditsDocs
SeedVR (image)varg.imageModel("seedvr")10Details
Topaz (image)varg.imageModel("topaz")15Details
SeedVR (video)varg.videoModel("seedvr-video")30Details
Topaz (video)varg.videoModel("topaz-video")50Details
Sima (video)varg.videoModel("sima-video-upscaler")15Details

​
Speech Models

ModelCodeCreditsDocs
Eleven V3varg.speechModel("eleven_v3")25Details
Eleven Multilingual V2varg.speechModel("eleven_multilingual_v2")25Details
Turbovarg.speechModel("turbo")20Details

​
Music

ModelCodeCreditsDocs
Music V1varg.musicModel()30Details

​
Transcription

ModelCodeCreditsDocs
Whispervarg.transcriptionModel("whisper")10Details
Groq Whisper Turbovarg.transcriptionModel("groq-whisper-large-v3-turbo")3Details

​
BYOK (Direct Provider Access)

If you have your own provider keys, you can use them directly: 
import { fal, elevenlabs, higgsfield, replicate } from "vargai/ai"

fal.imageModel("nano-banana-pro")
fal.videoModel("kling-v3")
elevenlabs.speechModel("eleven_v3")
elevenlabs.musicModel()
higgsfield.imageModel("soul")
Requires individual provider keys in .env. See BYOK for details.