Common Questions
Find answers about models, references, credits, API requests and content rights. Open the linked guide for the full steps and relevant policy.
How do I get started?
Open the site, sign in, select Start using, choose a task category and check the selected model price before sending a small request.
Where can I find my generated files?
Check Task list in the model workspace and Profile → My Works for generated media. The works page displays a 1–15 day retention notice, so save files locally promptly.
Which model should I choose?
Choose by output type, supported inputs, required controls and price. Use the Chat, Image, Video or Audio category, then compare the applicable service lines.
How do I write an image prompt?
State the subject, scene, composition, lighting and details to preserve. Set aspect ratio, resolution and output count in the model controls.
Why does the result ignore my prompt?
Compare the result with your requirements, simplify conflicting instructions and revise the specific detail that failed. Test the revision while keeping the other settings consistent.
Should I change one prompt detail at a time?
Yes. Changing one requirement at a time makes it easier to judge which revision helped. Keep the original prompt and settings for comparison.
How do I use a reference image?
Choose an image model with reference support. On the GPT Image 2.5 form, use + Reference image, then state what should stay the same and what should change.
Can I use more than one reference?
The image form supports multiple references. Check the current model and mode limits and give each reference a clear role.
Will the AI preserve the original image exactly?
Reference generation can redraw details. Compare the result with the source and preserve original artwork for text, logos or product details that must remain exact.
How do I turn a photo into a video?
Choose a video model with reference-image support, add the image, describe the intended motion and review duration, resolution and price before submission.
Which input image works best?
Use a clear image with a visible subject and room for the intended movement. Correct important face, hand, product or text details before animation.
Why did my generation fail?
Inspect the original task message and the corresponding usage record. The next action depends on whether the error concerns the input, a service failure or result display.
Should I retry or change my input?
Correct an identified input or parameter problem first. If the server outcome is unclear, check the existing task before submitting another paid request.
Will a failed generation use credits?
The site refund policy says charges for tasks that fail because of the site or upstream service are returned to account balance. A successfully produced result remains chargeable even when it does not meet your expectations.
Where is my result after a timeout?
Check the original Task list entry, Profile → My Works and Usage Records. Keep the task ID and investigate the original request before submitting a replacement.
How much will a generation cost?
Check the selected model, service line, billing unit and output settings in Pricing. Review the final consumption in Usage Records after the request.
Where can I check my credit usage?
Open More actions → Profile → Usage Records, choose AI App Platform or API Open Platform, then filter by date, model or task ID.
Do unused credits expire or roll over?
The site terms do not specify one universal expiry period for all credit types. Check the conditions of the specific purchase or promotion, or ask support about that balance type.
What is a token?
A token is a unit used to represent content for a model. It may correspond to a word, part of a word, punctuation or other encoded content; counting words does not give an exact token total.
Why do longer conversations cost more?
When previous conversation is included in a request, it adds input for the model to process. Longer inputs and replies can increase usage; the selected service determines cache and context handling.
When should I start a new chat?
Start a new chat for an unrelated task. For continuing work, carry forward a concise brief containing the relevant facts, decisions and constraints.
How do I fix a mispronounced name?
Put the difficult name in a short test passage, expand ambiguous abbreviations and adjust the spelling or supported pronunciation controls. Listen to that passage before generating a full narration.
Where do I get an API key?
Profile contains an API Keys entry for key management. Open it to review the current creation and permission controls; keep the resulting key in secure application storage.
Which base URL and model name should I use?
The site documentation shows https://api.vv666.ai with POST /v1/chat/completions for the selected Chat Completions model. Copy the exact model identifier from its current documentation.
Can I use an OpenAI-compatible client?
Use a compatible client for models whose site documentation lists the OpenAI-compatible protocol. Configure the client so the final endpoint contains /v1 exactly once, and check model-specific parameters.
Can I use generated content commercially?
Section 6 of the site terms permits commercial use within the license attached to the selected model or tool and subject to the conduct rules. Check that license for your intended use.
Who owns my uploaded and generated files?
The site terms say the platform does not claim ownership of your uploaded material, prompts or generated content, and receives permission to process it as needed to provide the service. Review the selected model license separately.
Are there model-specific restrictions?
Yes. The site terms defer the exact authorization scope to the selected model or tool. Obtain the applicable license before a commercial project if it is not displayed.
Are my uploads and chats private?
The policy describes collecting content for service delivery and forwarding necessary inputs to upstream model providers. Workspace Privacy mode masks account and balance information on screen; consult the policy for data processing.
Is my content used for model training?
The privacy policy does not give an explicit model-by-model no-training commitment. Obtain confirmation for the selected service before uploading material that requires that condition.
How do I delete my files or account?
The policy provides for deleting uploaded and generated material and requesting account closure. Use the account controls or current policy request instructions; backup removal and required record retention follow the policy.