📖 The AI Tool Bible

Lensa

Mobile photo editor with AI retouching, background removal, and the viral 'Magic Avatars' generator.

Freemium· Free download; Lensa Pro ~$35.99/yr; Magic Avatars sold as packsImage GenerationStable Diffusion (fine-tuned for Magic Avatars)6.8 / 10
Visit website →
Best for

Pick Lensa if you want a polished phone app for casual selfie retouching and the occasional batch of AI avatar portraits.

Skip if

Skip it if you need desktop control, an API, or pro-grade retouching for client work.

Lensa is Prisma Labs' consumer photo-editing app for iOS and Android, best known for its one-tap face retouching, background blur/replace, and the 'Magic Avatars' feature that fine-tunes a Stable Diffusion model on a handful of selfies to produce stylized AI portraits. The core editor leans on neural segmentation to split foreground from background so you can recolor, blur, or swap each independently, plus skin smoothing, eye brightening, and blemish removal that run on-device.

The target user is a phone-first creator who wants Instagram-ready selfies without learning Lightroom. Pricing is freemium: the app is a free download, but the useful retouching tools and Magic Avatars sit behind a Lensa Pro subscription (roughly $35.99/year, with Magic Avatar packs sold as one-off in-app purchases). It's a closed mobile product with no public API and no desktop version.

Lensa drew real controversy around the Magic Avatars launch over training-data sourcing and the way it handled NSFW outputs, so editorial buyers should weigh that alongside its slick UX. For pure casual portrait editing it's one of the better-polished mobile AI apps; for professional retouching or any kind of programmatic use, it's the wrong tool.

Editor's take

Lensa is the consumer face of Prisma Labs and one of the more competent mobile AI portrait apps, but it's a walled-garden app-store product, not a creator tool. The Magic Avatars trend has cooled and the subscription is steep for what's essentially a selfie filter pack.

— The AI Tool Bible editorial team

Pros

  • One-tap selfie retouching that genuinely looks clean on mobile
  • Strong foreground/background segmentation for portrait edits
  • Magic Avatars produce shareable stylized portraits from ~10-20 selfies
  • Available on both iOS and Android with a polished UI

Cons

  • ⚠️ Locked to mobile; no web app, desktop, or API
  • ⚠️ Most useful features gated behind a Pro subscription
  • ⚠️ Magic Avatars launch attracted criticism over training data and NSFW outputs
  • ⚠️ Limited control vs. dedicated tools like Lightroom or Midjourney

Use cases

selfie-retouchingai-avatarsbackground-removalportrait-filters

Explore related

Compare with similar tools

All in Image Generation

Midjourney

Featured
Image Generation · Midjourney v7
9.4

The gold standard for aesthetic AI image generation.

Paid· $10/mo Basic; up to $120/mo Megaillustrationconcept art

Flux

Featured
Image Generation · Flux.1 [schnell / dev / pro]
9.0

Black Forest Labs' open-weights image model — rivals Midjourney quality.

Freemium· API per-image; weights free for [schnell] and [dev]open sourceself-hosted

Stable Diffusion

Image Generation · SD 3.5 / SDXL
8.8

Open-source image generation — run anywhere, fine-tune anything.

Free· Free open weights; optional Stability APIlocalfine-tuning

Nano Banana (Gemini Image)

Image Generation · Gemini 3 Pro Image (Nano Banana Pro), Gemini 3.1 Flash Image (Nano Banana 2), Gemini 3.1 Flash-Lite Image (Nano Banana 2 Lite)
8.7

Google DeepMind's Gemini-powered image generation and conversational editing model family

Paid· Consumer access via Gemini app (free tier + Google AI Pro/Ultra subscriptions). API usage-based: Nano Banana Pro (Gemini 3 Pro Image) ~$0.134/image at 1K-2K, ~$0.24/image at 4K; Nano Banana 2 (Gemini 3.1 Flash Image) ~$0.067/image at 1K, up to ~$0.151 at higher resolutions; Nano Banana 2 Lite priced lower for high-throughput use. Batch API roughly 50% off. Enterprise pricing via Gemini Enterprise Agent Platform and Vertex AI.Marketing hero imagesProduct mockups and packaging visualisations

DALL·E 3

Image Generation · DALL·E 3
8.6

OpenAI's image model — strong on prompt adherence and text-in-image.

Freemium· Included in ChatGPT Plus; pay-per-image via APIpostersinfographics

Canva Magic Studio

Image Generation · Multi-model: partners including OpenAI (Magic Write historically on GPT models), Google Imagen and Runway for image/video, plus Canva's in-house design and layout models
8.5

Canva's all-in-one AI creative suite for design, image, video, copy, and presentations

Freemium· Free tier with limited Magic credits / Canva Pro ~$15/mo (or $120/yr) per user / Canva Teams ~$10/user/mo (3-seat min) / Canva Enterprise custom pricing. Pro unlocks higher Magic Write, Magic Media, and Magic Design usage caps.Social media post generationShort-form video ads