Paper.js vs Three.js

It is written in a language nearly identical to JavaScript, but adds helpful operator overloading to allow you to perform coordinate arithmetic.

An almost-unique feature: can test geometry if contains(), isInside() or intersects() one other object. Offers even hit-testing methods!

Objects in Paper make it easy to extend parent objects and run constructor functions without having to worry about JavaScript prototypal inheritance. Because of this it is easy to make compound drawing objects with their own instance variables and behaviors.

Example: each swimming tadpole follows its own behavior

These features make it easier to create objects that can act autonomously with complex behaviors. This makes Paper a good choice for particle effects and game development.

Elements all have the option of being created with a hash of properties which promotes clean, concise, and contextually local coding practice.

Paper uses a frame based approach to rendering the canvas. This makes it conceptually easier to create objects with animations and behaviors that interact and perform with other objects.

Along with Paper's approach to creating and managing objects this makes Paper especially good for creating applications with complex behaviors with many elements doing different actions at once, and makes it a good choice for making games.

Effects: Anaglyph, cross-eyed and parallax barrier.
Scenes: add and remove objects at run-time; fog
Cameras: perspective and orthographic; controllers: trackball, FPS, path and more
Animation: armatures, forward kinematics, inverse kinematics, morph and keyframe
Lights: ambient, direction, point and spot lights; shadows: cast and receive
Materials: Lambert, Phong, Standard, smooth shading, textures, PBR and more
Shaders: access to full OpenGL Shading Language (GLSL) capabilities: lens flare, depth pass and extensive post-processing library
Objects: meshes, particles, sprites, lines, ribbons, bones and more - all with Level of detail
Geometry: plane, cube, sphere, torus, 3D text and more; modifiers: lathe, extrude and tube
Data loaders: binary, image, JSON and scene
Utilities: full set of time and 3D math functions including frustum, matrix, quaternion, UVs and more
Export and import: utilities to create Three.js-compatible JSON files from within: Blender, openCTM, FBX, Max, and OBJ
Support: API documentation, public forum
Examples: Over 150 files of coding examples plus fonts, models, textures, sounds and other support files

The documentation is detailed, providing clear explanations and code samples of the various features. There are also hundreds of examples available.

Thanks to the recently introduced forum, it is easy to find a community of helpful developers.

With the introduction of the new MeshStandardMaterial, three.js support physically based rendering (PBR) out of the box allowing for real life quality material and lighting.

Three.js has importers for most of the industry standard files format (obj, mtl, fbx, 3ds, gltf, collada, babylon, playcanvas, stl, vrml, draco and many more), making it easy to author assets in your favourite modelling software and import them for use them in three.js.

Three.js has great project health, with activity on Github daily for bug fixes and new features.

Three.js offers a canvas renderer as a fallback when WebGL is not available.

Three.js support many special effects and post-processing filters including particles, lensflare, sprites, real time reflection and refraction and even area based lighting.

The core of the three.js system is kept to a minimum to reduce file size, however there are also hundreds of extensions maintained in the offical repo on github, along with many free textures, fonts and models. You can find them all here.

Three.js official documentation provides plenty of well-written examples with a wide variety of tutorials written by the community available that you can find by doing a google search.