Frequently asked questions
Straight answers about credits, rendering, characters and billing. Search it, or browse by topic.
DirecutAI is a hosted creative studio. Instead of juggling a dozen model providers, you get one interface, one credit balance and one asset library across every supported video and image engine — Veo, Seedance, Kling, Nano Banana, Seedream, GPT Image, Qwen and more.
No. Everything runs in the browser. Renders happen on our servers, so you can close the tab mid-generation and your results will be waiting in Assets when you come back.
Each preset appends a carefully tuned modifier to your prompt — lighting, lens, grade and motion language. Choosing None sends your prompt through completely untouched, which is the right choice when you have already written a very specific description.
Generations are priced in credits based on the selected AI model, resolution, duration and settings. The exact credit cost for your current configuration is always calculated dynamically and displayed directly on the Generate button before you run, so you always know your exact spend upfront.
You are refunded automatically. Credits are deducted when a job starts and returned in full if the provider reports a failure, returns no media, or the job times out. You never pay for a render you did not receive.
Subscription credits are granted for each billing period. Pay-as-you-go pack credits stay valid for 365 days from purchase. Your balance and plan status are always visible in your profile.
Card payments are processed by Garanti BBVA Virtual POS with 3D Secure authentication. Your card details are entered on the bank’s own page and never touch our servers.
Images typically land in seconds. Video ranges from under a minute to several minutes depending on the engine, resolution and duration. Everything continues server-side, and finished jobs appear in your Assets automatically.
Rendered media stays retrievable for 14 days from the generation date. Download anything you want to keep permanently — the Assets gallery gives you a direct link for every result.
Attaching one or more images switches the run into image-driven mode (I2V for video, I2I for images). Each engine accepts a different number of references — some take a single frame, some take a first and last frame, some take up to five. The prompt island shows the limit for whatever engine you have selected.
Yes — image engines support a batch of up to 4 in a single click. Each variation is charged as its own run, and the Generate button shows the combined total before you commit.
A character is a saved actor: a name, a description and a portrait. Type @Name in any Image or Video prompt and that character’s portrait is attached automatically as a reference image, which is what keeps their appearance consistent from shot to shot.
Yes. A tagged character counts as one of the engine’s reference images, exactly like a manual upload. If you tag more characters than the engine accepts, we tell you which one was not attached. Remove the tag from your prompt to free the slot again.
You can keep up to 20 characters at a time. Generating an AI portrait costs 1 credit; uploading your own photo is free. Portraits are compressed to 1024px on the longest edge and stored with your account, so they never expire the way rendered media does.
Open your profile from the sidebar or header, then use the camera button on your avatar. Images are cropped square and optimised to 512px automatically. You can remove the picture at any time to fall back to your initials.
Your live plan name appears under your name in the sidebar, in your profile, and at the top of the pricing modal — along with how many days remain before it renews. An expired subscription automatically reads as Free.