diff --git a/README.md b/README.md
index e629bfb..6432087 100644
--- a/README.md
+++ b/README.md
@@ -20,13 +20,13 @@ var myInt int
 f.Fuzz(&myInt) // myInt gets a random value.
 ```
 
-You can use it on string with different character, like ASCII, MultiByte, 
-CommonCJK, Chinese, Japanese, Number, Lower-Case, Upper-Case or 
-Random(each rune will be one of the above character).
+If you want to use it on string with different character like Chinese or Number, 
+you can get Unicode range from [Unicode](https://en.wikipedia.org/wiki/Plane_(Unicode)#Basic_Multilingual_Plane),
+and parse the low and high encoding into function `WithCharRange`. For example Chinese:
 ```go
 f := fuzz.New()
 var myString string
-f.WithStringGenMode(fuzz.Chinese).Fuzz(&myString)
+f.WithCharRange('\u4e00', '\u9fa5').Fuzz(&myString)
 ```
 
 You can use it on maps:
diff --git a/fuzz.go b/fuzz.go
index d724833..207641e 100644
--- a/fuzz.go
+++ b/fuzz.go
@@ -40,7 +40,7 @@ type Fuzzer struct {
 	maxElements       int
 	maxDepth          int
 	skipFieldPatterns []*regexp.Regexp
-	stringGenMode     int
+	characterRange    *charRange
 }
 
 // New returns a new Fuzzer. Customize your Fuzzer further by calling Funcs,
@@ -61,8 +61,8 @@ func NewWithSeed(seed int64) *Fuzzer {
 		minElements: 1,
 		maxElements: 10,
 		maxDepth:    100,
-		// random mode as default mode to gen string
-		stringGenMode: Random,
+		// characterRange will be nil if you not set the encoding range
+		characterRange: nil,
 	}
 	return f
 }
@@ -193,11 +193,12 @@ func (f *Fuzzer) SkipFieldsWithPattern(pattern *regexp.Regexp) *Fuzzer {
 	return f
 }
 
-// WithStringGenMode is used to set the string generation mode. Mode includes ASCII,
-// MultiByte, CommonCJK, Chinese, Japanese, Number, Lower case, Upper case or
-// random mode(each rune will use one character.).
-func (f *Fuzzer) WithStringGenMode(mode int) *Fuzzer {
-	f.stringGenMode = mode
+// WithCharRange is used to set the Unicode encoding range of the string.
+func (f *Fuzzer) WithCharRange(first, last rune) *Fuzzer {
+	f.characterRange = &charRange{
+		first: first,
+		last:  last,
+	}
 	return f
 }
 
@@ -285,9 +286,9 @@ func (fc *fuzzerContext) doFuzz(v reflect.Value, flags uint64) {
 
 	switch v.Kind() {
 	// String is common data struct but handle it with function randString
-	// randString should input string generation mode.
+	// randString should input character range although it is nil
 	case reflect.String:
-		v.SetString(randString(fc.fuzzer.r, fc.fuzzer.stringGenMode))
+		v.SetString(randString(fc.fuzzer.r, fc.fuzzer.characterRange))
 
 	case reflect.Map:
 		if fc.fuzzer.genShouldFill() {
@@ -445,13 +446,7 @@ func (c Continue) FuzzNoCustom(obj interface{}) {
 // RandString makes a random string up to 20 characters long. The returned string
 // may include a variety of (valid) UTF-8 encodings.
 func (c Continue) RandString() string {
-	return randString(c.Rand, c.fc.fuzzer.stringGenMode)
-}
-
-// RandChineseString makes a random string up to 20 characters long. The returned string
-// may include a variety of (valid) UTF-8 encodings. Rand unicode: \u4e00 - \u9fa5
-func (c Continue) RandChinesesString() string {
-	return randChineseString(c.Rand)
+	return randString(c.Rand, c.fc.fuzzer.characterRange)
 }
 
 // RandUint64 makes random 64 bit numbers.
@@ -538,49 +533,20 @@ var unicodeRanges = []charRange{
 	{' ', '~'},           // ASCII characters
 	{'\u00a0', '\u02af'}, // Multi-byte encoded characters
 	{'\u4e00', '\u9fff'}, // Common CJK (even longer encodings)
-	{'\u4e00', '\u9fa5'}, // Chinese
-	{'\u0800', '\u4e00'}, // Japanese
-	{'\u0030', '\u0039'}, // number
-	{'\u0061', '\u007a'}, // lower case
-	{'\u0041', '\u005a'}, // upper case
-}
-
-const (
-	ASCII = iota
-	MultiByte
-	CommonCJK
-	Chinese
-	Japanese
-	Number
-	LowerCase
-	UpperCase
-
-	Random = 10000
-)
-
-// randChineseString makes a random string up to 20 characters long. The returned string
-// may include a variety of (valid) UTF-8 encodings. Rand unicode: \u4e00 - \u9fa5
-func randChineseString(r *rand.Rand) string {
-	n := r.Intn(20)
-	sb := strings.Builder{}
-	sb.Grow(n)
-	for i := 0; i < n; i++ {
-		sb.WriteRune(unicodeRanges[3].choose(r))
-	}
-	return sb.String()
 }
 
 // randString makes a random string up to 20 characters long. The returned string
-// may include a variety of (valid) UTF-8 encodings.
-func randString(r *rand.Rand, genMode int) string {
+// may include a variety of (valid) UTF-8 encodings. If wants specific encoding
+// string, it should input character range.
+func randString(r *rand.Rand, characterRange *charRange) string {
 	n := r.Intn(20)
 	sb := strings.Builder{}
 	sb.Grow(n)
 	for i := 0; i < n; i++ {
-		if genMode == Random {
+		if characterRange == nil {
 			sb.WriteRune(unicodeRanges[r.Intn(len(unicodeRanges))].choose(r))
 		} else {
-			sb.WriteRune(unicodeRanges[genMode].choose(r))
+			sb.WriteRune(characterRange.choose(r))
 		}
 	}
 	return sb.String()