English
English

Google Flow Advanced Prompting

Google Flow Advanced Prompting

Get production-quality results from Google Flow with model-specific techniques for Veo 3.1 video and Nano Banana 2 images.

Updated in

Category

Video

From

Meowgic

Showcase

Description

Fix the problems generic prompting guides don’t cover. This advanced reference covers model-specific techniques for Google Flow’s Veo 3.1 video and Nano Banana 2 image models, from face drift and subject inconsistency to text artifacts, audio mismatch, and illustration-versus-photo errors.

What it covers

  • Reference image slots: 14 for Nano Banana 2 and Pro; Ingredients, First Frame, and Last Frame for Veo 3.1

  • Chaining outputs as inputs, scene by scene

  • Flow’s camera control vs. camera adjustment tools

  • Fixing subject drift, physics, audio, and negative prompts in Veo 3.1

  • First/last frame workflows with Nano Banana as the frame generator

  • Credit costs, Fast vs. Standard Veo, clip duration, seeds, and lip sync

  • A production assembly stack: backgrounds, color plates, realism, and multi-clip identity

Try asking

  • “Why does my Flow generation keep drifting the character’s face?”

  • “How do I get Nano Banana to make a photo instead of an illustration?”

  • “Set up a first and last frame transition video in Veo 3.1.”

What you get

Model-specific fixes and workflows for Flow, including which model to use for each task and when to spend credits, so your generations come out production-ready.

Information

Last updated

Runtime credits

Usage-based

Models

Auto

Give your AI team the skills to do more.

Give your AI team the skills to do more.

Give your AI team the skills to do more.

From product images to marketing campaigns, bring your everyday e-commerce tasks together with Meowgic.