Frequently Asked Questions

AIBase provides image and video generation in one workspace. You can create visuals from text, transform existing images, and produce short-form AI videos with selected production-grade models.

No. Beginners can start with plain-language prompts, while advanced users can add style, camera, and composition instructions for tighter control.

Use faster models for ideation and iteration, then switch to higher-quality models for final delivery. This workflow usually gives the best speed-to-quality ratio.

Credits are deducted per generation task based on model and mode. More advanced models and longer video durations generally consume more credits.

If a task fails before valid output is produced, the platform follows model-provider billing rules and returns status details for troubleshooting.

Yes. A common practice is to set internal generation guidelines by use case (draft, review, final) and assign model tiers per stage to control monthly spend.

Use specific prompts with clear subject, scene, style, and constraints. For image-to-image, upload clean references and align aspect ratio with your target output.

Model generation is probabilistic. Ambiguous prompts, conflicting instructions, or weak references can reduce consistency. Refine one variable at a time for stable improvements.

Temporary provider links may expire. AIBase is designed to persist generated assets through configured storage workflows, so production teams can keep deliverables accessible.

AIBase uses authenticated API routes and server-side provider calls, which helps avoid exposing provider keys in the browser.

Generated outputs can be stored in your configured object storage (for example Cloudflare R2), enabling persistent links and controlled delivery.