Models

What AI models TubeGen uses, and what each one costs

Quick answer

TubeGen uses Grok Imagine, GPT Image and Nano Banana 2 for image generation, and Seedance, Pixverse, Grok Imagine and Gemini Omni for animation. Scripts are written by TubeWriter, TubeGen's own writer, in four modes. Voiceover, soundtrack and avatars each run on a dedicated model. Every model is included in every plan and billed per generation in credits, from 22 credits for a Grok Imagine image to 120 for Nano Banana 2 at 1080p. Batch delivery cuts GPT Image and Nano Banana 2 costs by 40-50%.

Each stage of the pipeline uses the model suited to it. The models create the pieces; TubeGen writes the prompts, connects the stages, and assembles the video. For the wider category, see AI video generation models.

Credit costs

What each AI model costs in TubeGen

Every model is included on every plan. See plans and credits.

Image per image
Animation per second
Writing per 1,000 words
Voiceover
Soundtrack per second
Avatars per second

Batch delivery

GPT Image and Nano Banana 2 scenes can be returned within 24 hours for 40-50% less.

GPT Image · 720p
28 17 -40%
GPT Image · 1080p
62 37 -40%
Nano Banana 2 · 720p
80 40 -50%
Nano Banana 2 · 1080p
120 60 -50%

Comparison

Nano Banana 2 vs GPT Image vs Grok Imagine: which image model to pick

Grok Imagine

22
credits · 1080p
Best for
Photoreal, 3D-rendered
Resolutions
1080p
Consistent characters
1–2 per image, extras dropped
Choose it when
Cost matters most and the look is photoreal

GPT Image

28
credits · 720p
62
credits · 1080p
Best for
2D, illustrated
Resolutions
720p, 1080p
Consistent characters
Supported
Choose it when
Scenes are illustrated or text-heavy

Nano Banana 2

80
credits · 720p
120
credits · 1080p
Best for
Matching a saved art style
Resolutions
720p, 1080p
Consistent characters
Closest style adherence
Choose it when
Every video must match one channel look

Why more than one

Why TubeGen uses more than one image model

Image models have different strengths, so TubeGen offers three and lets you pick per project or per section. Animation works the same way. One model for everything would mean paying the top rate on scenes that do not need it.

The models vs the product

Is TubeGen just a wrapper around other AI models?

No. The models are components, and the product is the assembly.

What the models do

  • Write text from a prompt
  • Generate one image or one short clip from a prompt
  • Read a script aloud in a voice

What TubeGen does

  • Writes a specific prompt for every scene, in your art style and with your characters
  • Connects research, script, voice, visuals, music and edit in one place
  • Times visuals and music to the narration and assembles the cut
  • Adds its own tools: Niche Finder, AI Title Generator, Thumbnail Studio, Video Editor

If you want the outside view rather than ours, there is an honest TubeGen review and a list of TubeGen alternatives.

FAQs

Got Questions?
We've Got Answers.

Everything you need to know about the models behind TubeGen and what each one costs.

Not sure which model to pick?

Every model is on every plan. Compare credits on the pricing page or ask us in Discord.

See pricing

Pick the model. Build the video.

Every model on this page is included in every plan. The per-model credit costs above are all you pay per generation.