mj_turbo_reroll

Technical Specifications of mj-turbo-reroll

What is mj-turbo-reroll?

mj-turbo-reroll is CometAPI’s platform identifier for a Midjourney-compatible Turbo reroll capability. In practice, a “reroll” means regenerating an image job again so you get a new result set while keeping the original creative direction or task context. In Midjourney-compatible systems, reroll is treated as an image action alongside operations such as variation, outpaint, pan, and upscale-related actions.

The turbo part indicates that this model is mapped to Turbo-mode generation behavior. Midjourney documents Turbo Mode as a higher-speed GPU option available on newer Midjourney versions, with generation speeds that can be up to four times faster than Fast Mode.

Because reroll is an action on an existing job, mj-turbo-reroll is best understood not as a standalone text-to-image model for first submission, but as a specialized endpoint/workflow for regenerating prior image results quickly. Compatible APIs generally implement this by returning a task ID, then requiring you to fetch progress or receive a webhook callback when the rerolled task completes.

Main features of mj-turbo-reroll

• Turbo-speed regeneration: Designed for rerolling image jobs in Turbo mode, which upstream Midjourney documentation describes as significantly faster than standard Fast mode.
• Reroll-specific workflow: Focused on regenerating an existing task rather than creating an entirely new job from scratch. This is useful when you like the prompt direction but want different visual outcomes.
• Task-based asynchronous processing: Compatible Midjourney APIs typically return a task ID first, letting applications poll for completion or handle results asynchronously.
• Action/button integration: In many Midjourney proxy implementations, reroll is triggered via an action endpoint using a job-specific customId or button/action identifier extracted from a previous task result.
• Webhook-friendly architecture: Common Midjourney-compatible APIs support callback URLs so applications can receive job-status updates automatically instead of polling continuously.
• Fits multi-step image pipelines: Works well in production flows where users first imagine an image, inspect buttons/actions, then reroll, vary, pan, or upscale based on the returned task metadata.
• Midjourney-compatible semantics: Aligns with the broader Midjourney action ecosystem, where reroll sits alongside variation, outpaint, inpaint, and other post-generation operations.

How to access and integrate mj-turbo-reroll

Step 1: Sign Up for API Key

To access mj-turbo-reroll, first create an account on CometAPI and generate an API key from the dashboard. Store the key securely and load it through an environment variable in your application so it is not hard-coded in client-side code or public repositories.

Step 2: Send Requests to mj-turbo-reroll API

Use CometAPI’s standard API configuration and set the model field to mj-turbo-reroll. Because this model is used for a reroll workflow, your request will typically be part of a multi-step image pipeline in which you first create or fetch an existing Midjourney-style task, then submit the reroll action using the required task context.

curl https://api.cometapi.com/v1/responses \
  -H "Authorization: Bearer $COMETAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "mj-turbo-reroll",
    "input": {
      "task_id": "your_existing_task_id",
      "action": "reroll"
    }
  }'

Step 3: Retrieve and Verify Results

After submission, retrieve the task result using CometAPI’s response payload and job-tracking workflow. For mj-turbo-reroll, verification usually means confirming that a new task was created successfully, monitoring it until completion, and checking that the returned images correspond to a fresh reroll of the original task rather than the original result set.