YouTube Cover

From 5.3 credits

Generate scroll-stopping YouTube thumbnails from your video title, bold composition, expressive focal point, and a punchy headline rendered right on the image.

Video Title

Description(optional)

Style

Subject Image(optional)

Click to upload image

or drag and drop

Browse library

Still needed: Video Title

Runs on

1 models

Google
- Nano Banana Pro

Examples

Before / AfterNano Banana Pro

Bold: Pyra, fully automated

Gemini 3 Pro Image. Pyra's portrait becomes a billboard-bold YouTube thumbnail: face front and centre, a crisp baked "FULLY AUTOMATED" hook, her violet infinity mark as the symbol. One title in, a click-ready 16:9 cover out.

Reproduce

0103

YouTube Cover API

Call this pipeline from your own code. One request dispatches a run; the model is an input field, not a separate endpoint.

Get an API token

inputs: 4
required: 1
Pricing: 5.3 credits

Language

TypeScript Python Go CLI

bun add @pipe2-ai/sdk

Run a pipeline

import { createClient } from "@pipe2-ai/sdk";

const client = createClient(process.env.PIPE2_TOKEN!);

const { run_pipeline } = await client.RunPipeline({
  pipeline_slug: "youtube-cover",
  input: {
    title: "I Rebuilt My Studio in a Weekend",
  },
});

pip install pipe2

Run a pipeline

import asyncio
import os

from pipe2 import Pipe2Client

client = Pipe2Client(token=os.environ["PIPE2_TOKEN"])

run = asyncio.run(client.run_pipeline(
    pipeline_slug="youtube-cover",
    input={
        "title": "I Rebuilt My Studio in a Weekend",
    },
))

go get github.com/pipe2-ai/sdk-go

Run a pipeline

client := pipe2.NewClient(os.Getenv("PIPE2_TOKEN"))

input := json.RawMessage(`{
  "title": "I Rebuilt My Studio in a Weekend"
}`)

run, err := pipe2.RunPipeline(context.Background(), client, "youtube-cover", input)
if err != nil {
    log.Fatal(err)
}

brew install pipe2-ai/tap/pipe2

Run a pipeline

pipe2 pipelines run \
  --pipeline youtube-cover \
  --input-json '{"title": "I Rebuilt My Studio in a Weekend"}' \
  --wait

Per page

Input	Type	default	Description
`title`required	string	default—	The video title, drives the composition. Write it like a headline: specific, concrete, emotional.
`description`	string	default—	Optional extra context to shape the visual (channel genre, mood, key scene).
`style`	enum	defaultbold	Art direction preset. bold = saturated and dramatic, minimal = editorial and spacious, clickbait = extreme reaction face, cinematic = teal-orange film still, vlog = warm lifestyle, gaming = neon and motion blur. bold · minimal · clickbait +3 `boldminimalclickbaitcinematicvloggaming`
`subject_image`	string	default—	Optional: URL of a face or product photo to anchor the thumbnail.

Point any MCP client at pipe2 and your agent gets these tools. It finds this pipeline, reads the same schema above, then runs it.

list_pipelines
get_pipeline_schema
run_pipeline
get_pipeline_run_status
request_upload

https://mcp.pipe2.ai/mcp

{
  "mcpServers": {
    "pipe2ai": {
      "url": "https://mcp.pipe2.ai/mcp",
      "headers": { "Authorization": "Bearer YOUR_TOKEN" }
    }
  }
}

MCP setup

Full API reference CLI reference

Frequently Asked Questions

What size is the output?

16:9 at roughly 1920×1080 pixels, YouTube's recommended thumbnail resolution.

Can I overlay my own text?

Yes. The AI renders a short punchy hook headline (2 to 4 words distilled from your title) directly onto the image in the negative-space area. The thumbnail already ships with that baked-in text, and you can still add more of your own in any editor if you like.

Do I need to upload a photo?

No. If you don't provide a subject image, the AI invents a suitable scene from the title alone. A face or product photo anchors the result more reliably.

How many credits does it cost?

5.3 credits per thumbnail. Pricing is flat, re-run with a different title or style to A/B test.

Can I generate variations of the same thumbnail?

Yes, run the pipeline multiple times with slightly different titles or styles to A/B test which performs best.