AI models

Every major video, image and audio model on one credit wallet, and what each one is best at.

Try it with MIRO

Video

SeedanceByteDanceByteDance's video model for clean, cinematic motion at volume.
KlingKuaishouCamera control, motion transfer and video editing, up to 4K.
Veo 3.1GoogleGoogle's video model that generates sound with the picture.
WanAlibabaFast, low-cost drafts, video edits and transparent clips.
MiniMax H3 & HailuoMiniMaxExpressive faces and performances, plus inserting elements into footage.
LTX 2.3LightricksOpen video model with LoRA styles, audio-driven video and HDR.
Happy HorseAlibabaReference-driven video and edits for existing clips.

Image

FLUX.3Black Forest LabsPhotographic stills and controllable video from keyframes.
Nano BananaGoogleGoogle's image model for precise edits and consistent characters.
GPT ImageOpenAIOpenAI's image model for long briefs and legible text.
SeedreamByteDanceHigh-resolution key art, and images split into layers.
IdeogramIdeogramPosters and title cards with type that comes out right.
Recraft V4RecraftBrand graphics and editable vector art.
Qwen ImageAlibabaInstruction-following edits and layered images.

Audio

ElevenLabsElevenLabsNatural voiceover and full music tracks for your cut.
LyriaGoogleGoogle's music model for scores and background tracks.

Or tell MIRO what you want

MIRO is the AI agent inside G5 AI Studio. Describe the shot in plain language and MIRO picks the model, runs the generations and brings the results back into the chat.

Chat with MIRO