FASHN
Self-trained virtual try-on models, sold as an app and a credit-based API
Overview
FASHN is a virtual try-on and AI fashion-imagery platform that trains its own models, including the open-source FASHN VTON v1.5. It offers try-on, product-to-model, model swap, AI models, editing and short video through a web and mobile app on monthly plans and a credit-based developer API with Python and TypeScript SDKs.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 14 Oct 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
- API dashboard with request history
- Automated top-up alerts
- Commercial use
- 282 monthly credits
- Top-ups at $0.0675 per credit
- 4,150 monthly credits
- Top-ups at $0.06 per credit
- 25,594 monthly credits
- Top-ups at $0.0488 per credit
- 200 credits a month, rollover up to 600
- Up to 4K image generation and upscaling
- 720p AI video; 2 team members
- Up to 3 simultaneous generations
- 750 credits a month plus 50 a day
- 1080p AI video; 5 team members
- Up to 6 simultaneous generations
- Priority support
Use cases
What you can produce with FASHN
- Virtual try-on images of a garment on a person photo, about 5 to 17 seconds each
- Try-On Max outputs up to 4K, including shoes, hats, jewellery and bags
- On-model shots generated from product photos
- AI fashion models, model swaps and consistent face references
- Short videos generated from a single image
- Background removal and reframed images
- An open-source Apache 2.0 try-on model (FASHN VTON v1.5) to run yourself
ASEAN Perspective
FASHN in Southeast Asia
FASHN's API is cloud-hosted and globally accessible, making it technically available to ASEAN fashion e-commerce brands in Singapore, Malaysia, Indonesia, and Vietnam without any regional restrictions. The APAC fashion AI market is growing at over 40% annually and virtual try-on adoption is accelerating across platforms like Lazada and Shopee, creating natural demand for tools like FASHN. However, FASHN has not publicly documented ASEAN clients, local partnerships, or regional pricing in SGD/MYR/IDR, and its documentation and support appear primarily English-language. For ASEAN brands with development teams, the per-image API economics are competitive; for those without technical resources, the absence of a plug-and-play integration is a gap.
FASHN is a credible choice for virtual try-on because it builds its own models and documents them openly, from the Apache-licensed VTON v1.5 to clear API limits, retention rules and per-endpoint credit costs. Try-On Max reaches 4K and covers accessories, and failed predictions are free.
Two caveats: the app and the API are priced and credited separately, which makes comparing them awkward, and integration is developer work through the API and SDKs. Purchases are non-refundable, so use the ten free credits to test on your own catalogue first.
What we found
FASHN is a virtual try-on and fashion-imagery platform run by FASHN LTD, which gives a London registered office. Its about page says it was founded by Dan and Aya Bochman, a husband-and-wife team building independently since 2023 and still self-funded. Unlike many try-on tools that wrap a general image model, FASHN trains its own. Its research page describes FASHN VTON v1.5 as a 972-million-parameter model trained from scratch in pixel space on 18 million try-on pairs plus 4 million synthetic examples, released under Apache 2.0, which FASHN calls the first permissively licensed open-source try-on model. The open model outputs 576 by 864 pixels; the company is candid that this is lower than latent-space rivals.
The commercial products go further. Try-On v1.6 processes images at 864 by 1296 pixels in roughly 5 to 17 seconds depending on mode, and Try-On Max reaches 4K and handles shoes, hats, jewellery and bags. Other endpoints cover product-to-model, face-to-model, model creation and swap, editing, reframing, background removal and image-to-video, with Python and TypeScript SDKs and webhooks.
There are two separate price lists. The web and mobile app runs on subscriptions: Basic at $19 a month for 200 credits, Pro at $49 with 750 credits plus 50 a day, and Agency at $99 with 1,500 plus 100 a day and custom face references, with yearly billing worth about two months free. The API is on-demand at $0.075 a credit, minimum $7.50, or commitment tiers from $19 to $1,249 a month that cut top-ups to as little as $0.0488. A v1.6 try-on costs one credit; failed predictions cost nothing. App and API credits do not mix.
The operating details are documented: 50 requests a minute on the main endpoint, six concurrent jobs by default, and CDN outputs deleted after three days. New accounts get ten free credits and purchases are non-refundable, so test on those before committing.
Compiled by RECATOOLS from public sources: vendor documentation and pricing, code repositories and published reporting.
Notable facts
- FASHN was built by a husband-and-wife team working out of their living room in Belgrade, Serbia — inspired partly by watching the K-Drama 'Start-Up!' during the pandemic.
- The company remains fully bootstrapped and self-funded, reaching $25,000+ in monthly recurring revenue without venture capital.
- FASHN's v1.5 model operates in pixel space rather than latent space — an unusual architectural choice that preserves fine fabric texture and printed details more faithfully than VAE-based diffusion models.
- FASHN open-sourced its VTON v1.5 model weights on Hugging Face under Apache-2.0 in January 2026, allowing researchers to fine-tune the 972M-parameter model on consumer GPUs for as little as $5,000–$10,000.
Frequently asked questions
About this listing
This entry was compiled from publicly available data including FASHN's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with FASHN unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to FASHN directly →
Spotted something out of date? Suggest an update →
More in Image Generation
