diff --git a/README.md b/README.md
index 6b21fcd..6540712 100644
--- a/README.md
+++ b/README.md
@@ -26,10 +26,10 @@ const hexColor = buildRegex(
   capture(
     choiceOf(
       repeat({ count: 6 }, hexDigit),
-      repeat({ count: 3 }, hexDigit)
+      repeat({ count: 3 }, hexDigit),
     )
   ),
-  endOfString
+  endOfString,
 );
 ```
 
diff --git a/docs/API.md b/docs/API.md
new file mode 100644
index 0000000..e5c489f
--- /dev/null
+++ b/docs/API.md
@@ -0,0 +1,88 @@
+# API
+
+## Overview
+
+TS Regex Builder allows building readable regular expressions objects using easy-to-read domain-specific language (DSL).
+
+```ts
+// Same as: const regex = /Hello (\w+)/
+const regex = buildRegex(['Hello ', capture(oneOrMore(word))]);
+```
+
+This library exposes numerous regex components (e.g., `capture`, `oneOrMore`) corresponding to regular expression constructs. Components, when called, return `RegexElement` instances. You can form regular expressions by creating a sequence of elements and strings using JavaScript array literals:
+
+```ts
+const elements = ['Hello ', capture(oneOrMore(word))];
+```
+
+We refer to `RegexElement | string` type for brevity as `RegexNode`. Most of the regex components accept either a single regex node or an array of regex nodes:
+
+```ts
+oneOrMore('a'); // Single string node
+oneOrMore(anyOf('xyz')); // Single anyOf element
+oneOrMore(['a', anyOf('xyz')]); // Array of regex elements and strings
+```
+
+| Regex Component                    | Regex construct | Notes                                        |
+| ---------------------------------- | --------------- | -------------------------------------------- |
+| `buildRegex(...)`                  | `/.../`         | Create `RegExp`` object                      |
+| `anyOf('abc')`                     | `[abc]`         | Specify matching characters                  |
+| `characterRange('a', 'z')`         | `[a-z]`         | Specify range of matching characters         |
+| `capture(x)`                       | `(x)`           | Capture group                                |
+| `word`                             | `\w`            |
+| `digit`                            | `\d`            |
+| `any`                              | `.`             |
+| `whitespace`                       | `\s`            |
+| `startOfString`                    | `^`             | Start of string (or line in multiline mode)  |
+| `endOfString`                      | `$`             | End of string (or line in multiline mode)    |
+| `inverted(...)`                    | `[^...]`        | Inverts character class                      |
+| `characterClass(...)`              | `[...]`         | Construct concatenation of character classes |
+| `choiceOf(x, y, z)`                | `x\|y\|z`       | Alternative                                  |
+| `zeroOrMore(x)`                    | `x*`            |                                              |
+| `oneOrMore(x)`                     | `x+`            |                                              |
+| `optionally(x)`                    | `x?`            |                                              |
+| `repeat({ count: n }, ...)`        | `...{n}`        | Repeat exact number of times                 |
+| `repeat({ min: n, }, ...)`         | `...{n,}`       | Repeat at least given number of times        |
+| `repeat({ min: n, max: n2 }, ...)` | `...{n1,n2}`    | Repeat number of times in a range            |
+
+## Types
+
+```ts
+import type { EncodeOutput } from './encoder/types';
+
+export type RegexNode = RegexElement | string;
+
+export interface RegexElement {
+  type: string;
+  encode(): EncodeOutput;
+}
+```
+
+## `buildRegex`
+
+```ts
+function buildRegex(elements: RegexNode | RegexNode[]): RegExp;
+function buildRegex(
+  flags: {
+    global?: boolean; // Global search
+    ignoreCase?: boolean; // Case-insensitive search
+    multiline?: boolean; // Allows ^ and $ to match newline characters.
+    hasIndices?: boolean; // Generate indices for substring matches.
+    sticky?: boolean; // Perform a "sticky" search that matches starting at the current position in the target string.
+  },
+  elements: RegexNode | RegexNode[]
+): RegExp;
+```
+
+Parameters:
+
+- `elements` - single regex element or string or array of such elements
+- `flags` - flags for used for construction of `RegExp` object
+
+The top-level method for TS Regex Builder is `buildRegex`; it accepts either a single regex element, a string, or an array of such.
+
+## Character classes
+
+Te
+
+###