Skip to main content

How credits are charged in Agent One

Introduction

Credits in Agent One work a little differently from standard video creation. What you're charged depends on the models Agent One uses to generate your media and how much it creates in each exchange.

Every time Agent One responds to you, it may be doing several things at once, reasoning through your brief, generating images, producing video clips, designing voice, and more. Each of these actions draws from your credit balance based on the specific models running in the background.

The mode you choose directly affects how many credits are used per generation.

Lite vs Pro -> why it matters for credits

Agent One Pro: uses invideo's most powerful models - Claude Opus 4.6 for reasoning, Nano Banana Pro, GPT Image 2 or Seedream 5.0 as appropriate for image generations, Seedance 2.0 for video, Eleven labs Voice Design is default for Voice Design & ElevenLabs V3 for Text to Speech. Pro delivers the highest quality output but uses more credits per generation as a result.

Agent One Lite: uses faster, lighter models across the board - Grok 4.1 Fast for reasoning, Nano Banana 2 (1K) for images, Veo 3 fast, Pixverse V6, and Kling 3.0 is preferred, in that order, for video generations. For videos involving people, use Veo 3 fast. Pixverse is default for lipsync. Eleven labs V2 Multilingual is default for Text to Speech. Minimax Speech 2.6 is your default for Voice Design. When using premade characters, default to their fast model mix. Lite is more credit-efficient and ideal when you need to move quickly or are early in the creative process.

How to check credits used

After Agent One completes a response, click the three dots (···) on that message to see exactly how many credits were charged for that exchange, covering both the conversation and all media generated within it.

This gives you a clear view of where your credits are going, response by response, so you're never caught off guard.

💡 For a full breakdown of credit costs by model, resolution, and duration, see the video model pricing page.

Staying in control of generations

By default Agent One is set to Always allow, meaning it generates directly without asking for permission. If you'd prefer more control, you can change this by clicking the permission toggle at the top of the Agent One chat panel and selecting either:

  • Ask before generating videos - Agent One will ask for approval before generating video specifically

  • Always ask before generating - Agent One will ask for permission before any generation

When Agent One asks for approval you'll see a Generate Video popup showing exactly what it's about to create, the prompt, model, duration, and aspect ratio. You can approve as-is or adjust any of these before confirming.

This is especially useful when you want tighter control over credit usage, reviewing each generation before it runs means no surprises on your credit balance.

Did this answer your question?