Skip to content

Troubleshooting Guide

Solutions to common issues when using Get3W.

API Issues

401 Unauthorized

Symptoms: API returns 401 Unauthorized

Solutions:

  1. Verify your API key is correct and starts with sk_
  2. Check header format: Authorization: Bearer YOUR_API_KEY
  3. Check for extra spaces or line breaks introduced when copying the key
  4. Confirm the key hasn't been deleted, or generate a new one

A 403 Account suspended is different from 401 — it means the account tied to the key is disabled. Contact support in that case.

402 Payment Required

Symptoms: API returns 402 with an Insufficient balance message

Solutions:

  1. Check your balance with GET /v1/balance
  2. Top up — each task is priced before it runs and rejected if the balance cannot cover it
  3. Note that video and digital-human tasks cost far more than images
  4. If you're on trial credit, be aware some premium models require a paid balance

500 Internal Server Error

Symptoms: API returns 500 Internal Server Error

Solutions:

  1. Retry the request after a few seconds
  2. Check if the model is available
  3. Try a different model
  4. Contact support if the issue persists

Generation Issues

Task Stuck in "Processing"

Symptoms: Task status remains processing for too long

Expected wait times:

  • Images: < 30 seconds
  • Videos: < 5 minutes
  • 3D models: < 10 minutes

Solutions:

  1. Wait — queue time grows under load, visible as timings.queue_wait in the result
  2. Check the model's typical processing time, or use estimated_duration from the submit response
  3. Submit a fresh task rather than waiting indefinitely; there is no endpoint to cancel or delete a queued task
  4. Contact support with the task ID if it is consistently slow

Tasks are not stuck forever. A task that never starts is expired after 30 minutes, and one that stalls mid-run is expired after 60 minutes; either way it ends as failed and is not billed.

Poor Quality Output

Symptoms: Generated content doesn't match expectations

Solutions:

  1. Improve your prompt — see Write Better Prompts
  2. Try a different model — see Choose the Right Model
  3. Increase resolution or quality parameters
  4. Use the Prompt Enhancer in the web interface
  5. Try different seed values

Task Failed

Symptoms: Task status is failed with an error message

Common causes:

ErrorSolution
Invalid promptCheck prompt format and content
Invalid image URLEnsure the URL is accessible and valid
Unsupported dimensionsCheck supported sizes for the model
Content policy violationModify your prompt to comply with policy
Model unavailableTry later or use a different model

Integration Issues

Timeout Errors

Sync mode holds the connection open for the whole task, so slow models can outlive your HTTP client's default timeout. Either raise the timeout:

python
requests.post(url, headers=headers, json=payload, timeout=600)

Or, better for video and digital-human tasks, drop ?sync=true and poll instead — see Async Mode.

Web Interface Issues

Cannot Log In

  1. Clear browser cache and cookies
  2. Try a different browser
  3. Check if your account is active
  4. Contact support

Generation Not Starting

  1. Check your credit balance — the Run button shows the estimated cost before you click
  2. Try refreshing the page
  3. Check the browser console for errors

Still Need Help?

  • Check the FAQ for common questions
  • Contact support for personalized help

Released under the MIT License.