1
00:00:00,000 --> 00:00:01,978
How to Make an AI Audiobook From Your

2
00:00:02,018 --> 00:00:02,655
Manuscript.

3
00:00:04,255 --> 00:00:06,730
An AI audiobook is a book narrated by a

4
00:00:06,770 --> 00:00:09,307
synthetic voice instead of a person in a

5
00:00:09,347 --> 00:00:09,787
studio.

6
00:00:10,187 --> 00:00:12,991
With TTSLark you paste your manuscript,

7
00:00:13,031 --> 00:00:15,836
pick one voice and download the audio as

8
00:00:15,876 --> 00:00:18,373
an MP3 with matching subtitle files.

9
00:00:18,773 --> 00:00:21,007
It suits independent authors, serial

10
00:00:21,047 --> 00:00:23,580
fiction writers, story channels and anyone

11
00:00:23,620 --> 00:00:25,415
recording public domain books.

12
00:00:25,815 --> 00:00:28,069
There is no length limit on a single job,

13
00:00:28,109 --> 00:00:30,310
so a chapter or a whole book goes through

14
00:00:30,350 --> 00:00:31,044
in one pass.

15
00:00:33,644 --> 00:00:35,505
What do you need before you start an AI

16
00:00:35,545 --> 00:00:36,068
audiobook?

17
00:00:37,668 --> 00:00:39,847
You need a finished manuscript in plain

18
00:00:39,887 --> 00:00:42,177
text and the rights to publish it as audio

19
00:00:42,217 --> 00:00:42,272
.

20
00:00:42,672 --> 00:00:44,700
Your own book is fine, and so is a work

21
00:00:44,740 --> 00:00:46,617
in the public domain in the countries

22
00:00:46,657 --> 00:00:48,018
where you plan to sell it.

23
00:00:48,618 --> 00:00:50,886
Plain text matters more than most authors

24
00:00:50,926 --> 00:00:51,311
expect.

25
00:00:51,711 --> 00:00:54,070
A voice reads what is on the page, so

26
00:00:54,110 --> 00:00:56,716
running headers, page numbers and footnote

27
00:00:56,756 --> 00:00:59,239
markers copied from a PDF will end up in

28
00:00:59,279 --> 00:00:59,895
the audio.

29
00:01:00,295 --> 00:01:02,592
Open the manuscript in a text editor and

30
00:01:02,632 --> 00:01:03,856
clear those out first.

31
00:01:04,256 --> 00:01:06,368
Keep the chapter titles, since they are

32
00:01:06,408 --> 00:01:07,721
useful spoken signposts.

33
00:01:10,321 --> 00:01:12,471
How do you turn text into an audiobook,

34
00:01:12,511 --> 00:01:13,277
step by step?

35
00:01:14,877 --> 00:01:17,083
Turning text into an audiobook takes five

36
00:01:17,123 --> 00:01:19,222
steps, and most of the work sits in the

37
00:01:19,262 --> 00:01:19,850
first two.

38
00:01:20,450 --> 00:01:21,160
Step one.

39
00:01:21,560 --> 00:01:22,519
Prepare the text.

40
00:01:22,919 --> 00:01:24,944
Remove anything a listener should not

41
00:01:24,984 --> 00:01:27,221
hear and spell out abbreviations that read

42
00:01:27,261 --> 00:01:27,632
badly.

43
00:01:28,232 --> 00:01:28,938
Step two.

44
00:01:29,338 --> 00:01:30,264
Pick your narrator.

45
00:01:30,664 --> 00:01:32,406
Play the free samples in the voice

46
00:01:32,446 --> 00:01:34,535
library and choose one voice for the whole

47
00:01:34,575 --> 00:01:34,823
book.

48
00:01:35,223 --> 00:01:37,509
This guide is read by Maya, a voice we

49
00:01:37,549 --> 00:01:39,934
noted as warm and even over a long read.

50
00:01:40,534 --> 00:01:41,265
Step three.

51
00:01:41,665 --> 00:01:43,153
Render one chapter first.

52
00:01:43,553 --> 00:01:45,647
Listen to the full chapter, not just the

53
00:01:45,687 --> 00:01:46,093
opening.

54
00:01:46,493 --> 00:01:48,897
Names and unusual words show up here, and

55
00:01:48,937 --> 00:01:50,585
they cost little to fix now.

56
00:01:51,185 --> 00:01:51,912
Step four.

57
00:01:52,312 --> 00:01:53,587
Fix names and pacing.

58
00:01:53,987 --> 00:01:56,104
Add pronunciation and pause tags as

59
00:01:56,144 --> 00:01:58,436
covered below, then render that chapter

60
00:01:58,476 --> 00:01:58,884
again.

61
00:01:59,484 --> 00:02:00,262
Step five.

62
00:02:00,662 --> 00:02:01,528
Render the book.

63
00:02:01,928 --> 00:02:03,582
Paste the rest chapter by chapter.

64
00:02:03,982 --> 00:02:06,212
Rendering runs on our servers, so you can

65
00:02:06,252 --> 00:02:08,101
close the tab and come back later.

66
00:02:08,700 --> 00:02:10,797
The separate files make later edits cheap.

67
00:02:11,197 --> 00:02:13,654
If a reader spots a typo in chapter 12,

68
00:02:13,694 --> 00:02:16,212
you render chapter 12 again and leave the

69
00:02:16,252 --> 00:02:16,983
rest alone.

70
00:02:19,583 --> 00:02:21,896
How should you format a manuscript for an

71
00:02:21,936 --> 00:02:23,280
AI audiobook generator?

72
00:02:24,880 --> 00:02:27,074
Format the manuscript the way you want it

73
00:02:27,114 --> 00:02:27,486
spoken.

74
00:02:27,886 --> 00:02:30,305
An AI audiobook generator follows your

75
00:02:30,345 --> 00:02:32,948
punctuation closely, so commas, full stops

76
00:02:32,988 --> 00:02:35,530
and paragraph breaks become the rhythm of

77
00:02:35,570 --> 00:02:36,492
the narration.

78
00:02:37,092 --> 00:02:37,823
Paragraphs.

79
00:02:38,223 --> 00:02:40,117
Leave a blank line between them, which

80
00:02:40,157 --> 00:02:41,753
gives the voice a natural break.

81
00:02:42,353 --> 00:02:43,368
Chapter openings.

82
00:02:43,768 --> 00:02:45,820
Write the title as a sentence, such as "

83
00:02:45,860 --> 00:02:46,626
Chapter Twelve.

84
00:02:47,026 --> 00:02:48,156
The Long Road North."

85
00:02:48,556 --> 00:02:50,422
Then add a two second pause before the

86
00:02:50,462 --> 00:02:51,033
first line.

87
00:02:51,633 --> 00:02:52,737
Numbers and dates.

88
00:02:53,137 --> 00:02:54,926
Spell them out when the reading matters.

89
00:02:55,326 --> 00:02:57,866
"Nineteen eighty four" is safer than a

90
00:02:57,906 --> 00:03:00,034
bare 1984 in a historical novel.

91
00:03:00,634 --> 00:03:01,277
Dialogue.

92
00:03:01,677 --> 00:03:03,007
Keep attribution simple.

93
00:03:03,407 --> 00:03:05,726
"She said" reads cleanly, and a long tag

94
00:03:05,766 --> 00:03:07,227
can slow a fast exchange.

95
00:03:07,827 --> 00:03:09,090
Front and back matter.

96
00:03:09,490 --> 00:03:11,596
Add a short spoken credit at the start

97
00:03:11,636 --> 00:03:13,795
and leave out the index and the table of

98
00:03:13,835 --> 00:03:14,371
contents.

99
00:03:14,971 --> 00:03:17,105
Text to audiobook work rewards a slow

100
00:03:17,145 --> 00:03:17,945
first chapter.

101
00:03:18,345 --> 00:03:20,501
Whatever you learn there applies to all

102
00:03:20,541 --> 00:03:22,803
the others, so a careful hour at the start

103
00:03:22,843 --> 00:03:25,092
can save a full day of corrections later.

104
00:03:27,692 --> 00:03:29,899
How do you keep AI audiobook narration

105
00:03:29,939 --> 00:03:31,553
consistent across chapters?

106
00:03:33,153 --> 00:03:35,449
Use the same voice and the same speed for

107
00:03:35,489 --> 00:03:37,784
every chapter, and write down both before

108
00:03:37,824 --> 00:03:38,380
you begin.

109
00:03:38,780 --> 00:03:40,927
A long book may take several sittings,

110
00:03:40,967 --> 00:03:43,223
and a small speed change between chapters

111
00:03:43,263 --> 00:03:44,193
is easy to hear.

112
00:03:44,793 --> 00:03:46,528
Character names need the same care.

113
00:03:46,928 --> 00:03:49,056
If a name comes out wrong, select it in

114
00:03:49,096 --> 00:03:51,171
the script and use Pronounce as to give

115
00:03:51,211 --> 00:03:52,744
the spelling you want spoken.

116
00:03:53,144 --> 00:03:55,063
The tag holds the spelling you want

117
00:03:55,103 --> 00:03:57,341
spoken, and the written name stays in your

118
00:03:57,381 --> 00:03:57,964
subtitles.

119
00:03:58,564 --> 00:04:00,760
Keep a short list of these tags and paste

120
00:04:00,800 --> 00:04:02,784
them into each new chapter, so a name

121
00:04:02,824 --> 00:04:04,954
never changes halfway through the book.

122
00:04:05,554 --> 00:04:06,948
Pacing is the other tool.

123
00:04:07,348 --> 00:04:09,599
Add a two second pause between scenes and

124
00:04:09,639 --> 00:04:11,891
a half second pause before a line you want

125
00:04:11,931 --> 00:04:12,358
to land.

126
00:04:12,758 --> 00:04:15,037
To stretch a single phrase, select it and

127
00:04:15,077 --> 00:04:15,886
use Slow down.

128
00:04:16,486 --> 00:04:18,855
If no stock voice feels right, blend two

129
00:04:18,895 --> 00:04:21,323
voices with Voice Mix and save the result

130
00:04:21,363 --> 00:04:22,538
under your own name.

131
00:04:22,938 --> 00:04:25,152
The blend stays available for every later

132
00:04:25,192 --> 00:04:26,136
book in a series.

133
00:04:28,736 --> 00:04:30,511
What does an AI audiobook cost?

134
00:04:32,111 --> 00:04:34,930
An AI audiobook costs one credit for every

135
00:04:34,970 --> 00:04:36,367
10,000 bytes of text.

136
00:04:36,767 --> 00:04:39,036
The generator shows the exact cost before

137
00:04:39,076 --> 00:04:39,613
you start.

138
00:04:40,013 --> 00:04:42,327
English runs close to one byte per

139
00:04:42,367 --> 00:04:45,007
character, so a 100,000 word novel comes

140
00:04:45,047 --> 00:04:47,491
to roughly 60 credits with a standard

141
00:04:47,531 --> 00:04:47,923
voice.

142
00:04:48,323 --> 00:04:50,670
That fits inside one month of the Starter

143
00:04:50,710 --> 00:04:53,057
plan at six dollars and ninety nine cents,

144
00:04:53,097 --> 00:04:54,429
with credits left over.

145
00:04:55,029 --> 00:04:57,384
The free plan gives you five credits that

146
00:04:57,424 --> 00:04:59,836
never expire, enough to test your narrator

147
00:04:59,876 --> 00:05:01,946
on a full chapter before you pay for

148
00:05:01,986 --> 00:05:02,500
anything.

149
00:05:02,900 --> 00:05:04,932
Our audio length calculator estimates

150
00:05:04,972 --> 00:05:07,111
running time from your word count, which

151
00:05:07,151 --> 00:05:08,746
helps when you plan chapters.

152
00:05:11,346 --> 00:05:13,456
Does Audible accept AI generated

153
00:05:13,496 --> 00:05:14,278
audiobooks?

154
00:05:15,878 --> 00:05:17,804
Not through a standard ACX upload.

155
00:05:18,204 --> 00:05:21,483
The April 15, 2026 version of ACX's audio

156
00:05:21,523 --> 00:05:24,724
submission requirements says an audiobook

157
00:05:24,764 --> 00:05:27,425
must be narrated by a human unless

158
00:05:27,465 --> 00:05:29,086
otherwise authorized.

159
00:05:29,486 --> 00:05:31,623
It also prohibits unauthorized text to

160
00:05:31,663 --> 00:05:32,044
speech.

161
00:05:32,444 --> 00:05:34,808
The same page says Audible is working to

162
00:05:34,848 --> 00:05:37,041
accept third party text to speech from

163
00:05:37,081 --> 00:05:39,388
publishers, so check it again before you

164
00:05:39,428 --> 00:05:40,344
plan a release.

165
00:05:42,944 --> 00:05:44,851
Where can you publish an AI audiobook?

166
00:05:46,451 --> 00:05:48,856
You can publish an AI audiobook on Google

167
00:05:48,896 --> 00:05:50,700
Play Books or sell it directly.

168
00:05:51,100 --> 00:05:53,534
In Google's Partner Center you choose

169
00:05:53,574 --> 00:05:56,009
Synthesized voice as the narrator type

170
00:05:56,049 --> 00:05:58,484
when you upload, and MP3 is one of the

171
00:05:58,524 --> 00:05:59,602
accepted formats.

172
00:06:00,002 --> 00:06:02,384
Selling directly from your own site or a

173
00:06:02,424 --> 00:06:04,632
storefront like Gumroad gives you full

174
00:06:04,672 --> 00:06:07,112
control over price and format, and a story

175
00:06:07,152 --> 00:06:09,418
channel can publish chapters on YouTube

176
00:06:09,458 --> 00:06:11,418
with the subtitle files attached.

177
00:06:12,018 --> 00:06:14,015
Store rules change often, so read the

178
00:06:14,055 --> 00:06:16,266
current terms of each store before release

179
00:06:16,306 --> 00:06:16,360
.

180
00:06:16,760 --> 00:06:18,893
Keep your own master files too, because a

181
00:06:18,933 --> 00:06:21,056
new store may ask for a different format.

182
00:06:23,656 --> 00:06:25,070
Frequently asked questions.

183
00:06:27,670 --> 00:06:30,175
Can I use AI to generate an audiobook for

184
00:06:30,215 --> 00:06:30,579
free?

185
00:06:31,179 --> 00:06:32,419
Yes, for short works.

186
00:06:32,819 --> 00:06:35,335
The free plan includes five credits that

187
00:06:35,375 --> 00:06:37,708
never expire, each up to 10 minutes of

188
00:06:37,748 --> 00:06:40,243
audio, with subtitles and commercial use.

189
00:06:40,643 --> 00:06:42,299
A short story fits comfortably.

190
00:06:42,699 --> 00:06:44,750
A novel needs a paid plan or added

191
00:06:44,790 --> 00:06:45,313
credits.

192
00:06:46,913 --> 00:06:49,397
How long can one AI audiobook render be?

193
00:06:49,997 --> 00:06:51,313
There is no fixed limit.

194
00:06:51,713 --> 00:06:53,942
Your text is split on our servers and

195
00:06:53,982 --> 00:06:56,443
comes back as one continuous file with one

196
00:06:56,483 --> 00:06:56,832
voice.

197
00:06:57,232 --> 00:06:59,409
We still suggest one render per chapter,

198
00:06:59,449 --> 00:07:01,401
because it makes corrections easier.

199
00:07:03,001 --> 00:07:04,545
Can I use more than one voice for

200
00:07:04,585 --> 00:07:05,609
different characters?

201
00:07:06,209 --> 00:07:08,667
Each render uses one voice, which is what

202
00:07:08,707 --> 00:07:10,809
most listeners expect from a single

203
00:07:10,849 --> 00:07:11,384
narrator.

204
00:07:11,784 --> 00:07:13,734
For a two narrator book, render each

205
00:07:13,774 --> 00:07:16,025
narrator's chapters with a separate voice.

206
00:07:17,625 --> 00:07:19,914
Do I own the audio from my AI audiobook?

207
00:07:20,514 --> 00:07:23,100
Everything you render carries a commercial

208
00:07:23,140 --> 00:07:25,604
license on every plan, with no watermark

209
00:07:25,644 --> 00:07:27,354
and no attribution required.

210
00:07:27,754 --> 00:07:29,371
You still need the rights to the book

211
00:07:29,411 --> 00:07:29,750
itself.

212
00:07:32,350 --> 00:07:33,562
Start with one chapter.

213
00:07:35,162 --> 00:07:37,288
Paste your opening chapter into the text

214
00:07:37,328 --> 00:07:39,401
to speech generator, choose a voice and

215
00:07:39,441 --> 00:07:40,815
listen to the whole thing.

216
00:07:41,215 --> 00:07:43,614
If you want a narrator who fits a darker

217
00:07:43,654 --> 00:07:45,995
or more dramatic book, the narrator text

218
00:07:46,035 --> 00:07:48,125
to speech page is a good next stop.
