Welcome! Find short, expandable answers about setup, workflow, quality, legal, and API.
What is Typelense?
Typelense is an AI product photography platform for fashion & e-commerce. Upload garments, choose a model & pose, optionally a background—and get consistent, realistic product images for shop, social, and ads.
Who is Typelense for?
For D2C brands, marketplace sellers, agencies, and studios that need large volumes of images produced quickly, consistently, and cost-effectively—including campaign visuals, on-model, ghost-mannequin, flat-lay upgrades, and creative variations.
What does a typical generation look like?
Upload garments (top/bottom/set)
Select a model (from gallery or your brand model)
Lock the pose (pose remains 1:1 unchanged)
Choose background (studio, location, gradient, or your own upload)
Render & review variants (batch/series supported)
Will the pose really remain unchanged?
Yes. The selected pose is kept 1:1. Only clothing and background change—ideal for catalog comparability.
Which file formats & quality should I upload?
PNG or JPG, Webp. Recommended: clean cutouts or clear flat-lay/on-hanger photos, one garment view per file, minimal wrinkles/reflections, no collages. Higher resolution ⇒ better edges and detail.
What aspect ratios & resolutions do I get?
Common ratios: 1:1, 4:5, 3:4, 2:3, 9:16—ideal for shops, social, and ads. Optional upscale for high-resolution exports. (Project-wide defaults are configurable.)
How accurate is garment fidelity (logos, patterns, fabrics)?
Very high—especially with clean input photos. Best results with even lighting, visible texture, and minimal compression artifacts. Extremely fine patterns (moiré), highly reflective, or transparent fabrics are inherently more challenging.
Is underwear/swimwear allowed?
Yes, provided it’s non-explicit and focused on product depiction. No minors, no sexualized content. We reserve the right to moderate violations of our guidelines.
What about watermarks/AI markings?
Typelense does not add visible watermarks. Some underlying models may embed invisible watermarks/signatures. We do not remove or tamper with such signals.
Who owns the generated images?
Generally, you do, provided you hold the rights to the inputs (garments, references, logos). You receive commercial usage rights to the generated assets. Please also respect the license terms of any model/asset sources used in your project.
Where is data processed & how long is it retained?
Stored only as long as needed for the service (render, review, re-use) or per contract. You can delete assets/jobs at any time. Enterprise: options for shortened retention, customer-specific data residency, and DPA/AVV.
How are credits/pricing calculated?
Billed per generation/variant (and optional upscale). Failed jobs are not charged. Volume plans and enterprise packages available.
What can I do if quality/fit isn't right?
Improve input: higher resolution, one garment per file, clean edges
Check pose: avoid heavy self-occlusions; keep pose 1:1 fixed
Simplify background: use studio/gradient for catalog sets
More variants: 4–8 variants can improve hit rate
If you’re stuck, send us the job ID & example—we’ll debug it with you.
How do I ensure series consistency (catalog look)?
Define project presets (pose set, lighting, background, ratio) and use uniform inputs per category. Repeat presets across all products in the series.
Which compliance topics do you cover (EU AI Act, data protection)?
We operate with privacy-by-design, provide DPA/AVV, role-based access control, and content moderation. Enterprise: additional controls (e.g., audit logs, data residency, pen-tests) on request.
Do you offer enterprise features (SLA, SSO, roles)?
Yes. SLA, SSO (SAML/OIDC), fine-grained roles/permissions, dedicated environments, custom quotas, and support channels are available.
How do I reach support?
In-product via Help → Support, by email at [email protected]or via a shared Slack channel (Enterprise). Please include your project/job ID.