PIXLR

Pixlr Editor gives you layers, blend modes, masks, adjustment layers, and a history panel without an install. It covers the retouching and compositing work that sends most people to a desktop app, and it opens PSD files. The interface will feel familiar if you have used Photoshop, minus the deepest color and automation tools.

The generator pulls from several external models, including Flux, Recraft, SDXL, Qwen, and Stable Diffusion, instead of locking you into one. You can generate from a text prompt alone or use a reference image for style transfer. Model choice matters because each one draws differently and each one costs a different number of credits.

Two separate tools handle cleanup. The background remover isolates a subject in one click and exports a transparent PNG. The object remover works like a brush: you paint over a stranger in a holiday photo or a cable across a floor, and the area gets rebuilt from surrounding pixels. Both run in seconds.

Generative fill lets you select a region and describe what should appear there. Generative expand grows the canvas in any direction and fills the new space. Generative transform moves, resizes, or rotates an object and rebuilds whatever it uncovers. Together they turn a cropped photo into something you can reframe for a different format.

Pixlr Agent chains generation, editing, and effects into a repeatable workflow instead of a single prompt. You pick a template for social posts, branding, or product shots, or build a session from scratch. It is the closest thing here to an automation layer, and it saves the manual step hopping on recurring jobs.

The video generator turns text or a still image into a short clip using models such as Kling, Veo, and Seedance. The audio tools cover speech generation and music from providers like ElevenLabs. Both live in the same credit pool as image generation, so a heavy video day eats the allowance that would have gone to images.